The International Criminal Court's annual budget sits just under €200 million. It pays judges, investigators, forensic accountants, interpreters and victim liaisons out of accounts in The Hague. Almost none of that money is denominated in dollars. Almost all of it eventually clears through New York.
That is the entire story once you strip away the politics. When the Trump administration escalated from visa bans to financial designations against the court — an executive order in February, judge-level designations following through the spring and summer — it did not attack the ICC's jurisdiction. It did not invalidate a single warrant. It attacked the court's payment graph. And the payment graph is where crypto lives now.
Three years ago that distinction mattered far less. The court's bankers were European, its donors were sovereign states, and Washington could apply friction only at the edges. The sanctions toolkit has changed shape. Treasury's designation machinery has been pointed at smart contracts, at exchanges, at mixers, at stablecoin issuers who clear more offshore dollar volume than most mid-sized banks. A court that was never designed to have a crypto exposure suddenly has one, and nobody has modeled what happens next.
The ICC was created by the Rome Statute and funded by 125 states parties through assessed contributions plus voluntary top-ups. Roughly two-thirds of the budget is assessed. The rest is voluntary, and the voluntary bucket is where the interesting money moves — Trust Fund for Victims programs, field operations, forensic contractors, and, over the last two years, a genuinely serious digital-asset seizure capability inside the Office of the Prosecutor.
That capability is the part nobody outside The Hague discusses. The OTP does not merely prosecute; it recovers. When defendants flee with assets — real estate, aircraft, and increasingly, wallet balances — the prosecutor's office pursues forfeiture. Digital assets seized in that process require custody. Custody requires a regulated provider. Regulated providers, at the top of the food chain, require a US correspondent relationship.
You can see the problem forming from three moves away.
As for why the response took a financial rather than diplomatic form: the ICC issued arrest warrants touching US and Israeli personnel, and Washington reached for the tool that actually has teeth in 2026. Visa bans are symbolic. Designations are operational. An SDN entry does not argue with you. It tells every bank, exchange, payment processor, custodian and stablecoin issuer inside the dollar system that touching you is itself a violation.
There is a precedent worth keeping in view. OFAC's 2022 designation of Tornado Cash's smart contracts — and its 2025 delisting — demonstrated that the US is willing to sanction code, and that undoing a code sanction is far messier, legally, than imposing one. The ICC case runs a version of the same experiment in the opposite direction. Instead of sanctioning an anonymous protocol, it sanctions a named international institution with staff, payroll, a bank account and a legal department. That sounds easier. It isn't. Institutions have vendors, and vendors have vendors.
The first hop is boring and decisive. The ICC's operating accounts sit with European banks — institutions that are not in New York and do not need to be. Dollar clearing runs through correspondent relationships, and every large European bank with an appetite for dollar business has a US nexus Washington can threaten. When a designation lands on a client, the bank's choice is not between the client and the regulator. It is between the client and its own dollar access. That choice has one answer, and it has had the same answer for forty years.
The second hop is the one that will be litigated, and it is the sleeper in this entire affair. The European Union maintains a Blocking Statute — Council Regulation 2271/96 — that prohibits EU operators from complying with specified US secondary sanctions and hands them a cause of action for damages if they do. Its annex list is a relic of the 1990s Cuba and Iran fights, and it has never been seriously stress-tested against a designation aimed at a body the EU itself hosts on its own soil. The Netherlands is a member of the ICC. Dutch banks are EU operators. If a designation touches the court's banking relationships, a compliance officer in Amsterdam faces a US directive to block and an EU regulation instructing them not to comply. That contradiction is the actual battlefield, and it has nothing to do with the merits of any warrant.
The third hop is crypto, and it compresses the timeline from months into seconds.
Stablecoin issuers are the choke point. Tether and Circle are centralized by design; their contracts contain freeze and blacklist functions that respond to law-enforcement and OFAC requests. The operational record is not ambiguous. Tether has frozen billions cumulatively across sanctioned and flagged addresses. Circle's USDC contract does the same, under the same pressure, with the same legal reasoning. There is no governance vote, no validator set to convince, no fork to threaten. There is a function call and a transaction, and the latency is measured in minutes. If ICC-linked addresses ever appear on a designation, the freeze is not a risk to be assessed. It is a schedule to be published.
Composability isn't insulation from a freeze order — it's an amplifier, because every contract downstream of a frozen address inherits the contamination. A vendor's multisig treasury, an escrow contract, an OTC desk that settled the liquidation of seized assets: none of them need to be named. They only need to be reachable. This is the part of the modular thesis that builders consistently underestimate — the same property that makes a protocol useful to strangers makes it useful to whoever is screening the strangers.
I have watched this perimeter collapse in real time before. In October 2017, I spent 48 hours cross-referencing Parity's Rust source against Etherscan logs while the community argued in public about what had actually broken, and published the root cause within four hours of the fork announcement. The lesson was never about Parity. It was that institutional assumptions about which parties can be isolated are usually wrong by one hop. The ICC has more hops than a wallet. That means more nodes to fail, and a longer window in which failing nodes are invisible.
The evidence-custody problem is the detail I would flag first to any compliance officer evaluating exposure here, and it has gone almost entirely unreported. Prosecutors seize crypto from fugitives. Assets seized that way have to sit somewhere — a qualified custodian, a qualified hardware arrangement, or an exchange account. Based on my audit experience reviewing custody and storage arrangements across institutional counterparties, the failure mode is nearly always the same: the arrangement is resilient along the axis it was designed for and brittle along the one it inherited from a banking partner. If the ICC is designated and its custody provider is US-regulated or has US-regulated affiliates, that provider is holding assets on behalf of a designated entity. Freezing the evidence is not a legal edge case. It is the straightforward reading of the 50% rule and secondary sanctions exposure. The sanction would not merely cut the court off from money. It would freeze the court's ability to hold the proceeds of the crimes it exists to prosecute.
I ran a related experiment in early 2026: five autonomous trading agents on a testnet, each with signing authority, each fed adversarial inputs, each monitored for prompt-injection drains. The finding that unsettled institutional readers was not the exploit itself. It was how quickly automated compliance logic inherited the ambiguity of its own prompt. Screening systems are increasingly model-driven. Feed a model a designation with a fuzzy perimeter — "associated with," "acting for or on behalf of" — and you get a distribution of freeze decisions across exchanges and custodians rather than a uniform rule. That distribution is where real people lose access to real money, and nobody publishes a list of who got caught inside it.
Which brings us to the mechanism that actually determines the outcome, and the one that never appears in a press release.
The blast radius of a designation is set by pre-emptive de-risking, not by the designation itself. Banks and exchanges do not wait for the SDN entry to publish. They de-risk on rumor, on executive orders, on reporting that sanctions are being prepared. Legal teams ask the only question that matters: do we want to be the institution that processed this transaction in six months, in front of a congressional committee? The answer is almost always no, and it is almost always given before anyone has read the statute.
The compliance desk can't wait. It does not have the information required to wait. Pre-clearance is slow, the reputational cost of a false positive is asymmetric, and the cost of a false negative is existential. When I modeled the TerraUSD death spiral in May 2022 with three independent developers, the part that transferred to this problem was not the liquidity math — it was the reflexivity. Redemptions beget redemptions. De-risking begets de-risking. The formal scope of a sanctions program systematically understates its footprint, and in the ICC's case the gap is measured in vendors, contractors and researchers who have never held a dollar account in New York.
A further wrinkle should worry anyone assuming this stays confined to a court in the Netherlands. Institutional sanctions create compliance gray zones that propagate outward. A foundation that funds legal aid. A university holding a research grant. A DAO treasury that donated to a humanitarian program the court administers. None of these are obvious targets, and all of them are reachable through the "acting on behalf of" language that gives OFAC its flexibility. They will be screened, de-risked and quietly cut loose long before any attorney drafts a memo, and most of them will never learn why.
There is an ironic second-order effect worth naming. If the court wants to keep paying people, the venues still willing to serve it are the ones already comfortable with compliance risk — counterparties with thinner controls, worse documentation and less willingness to ask questions. A designation intended to isolate an institution instead pushes it toward the least transparent segment of the market. That is not a prediction about the ICC's intent. It is a structural property of how financial exclusion works, and it has played out identically in every sanctions regime I have covered since 2017.
Quantify the stakes honestly. The court's assessed contributions run into the hundreds of millions of euros annually across 125 states parties. Nobody needs to move all of it. A credible non-dollar channel handling a fifth of the operating budget would be roughly €40 million a year — rounding error against global FX volumes, and completely irrelevant to the argument. The value is not in the flow. The value is in the proof.
The claim that code is neutral stopped being neutral in 2022, and anyone still clinging to it is defending something that never existed. But the correction on offer — build an identity layer that permanently brands every participant — is worse. The reason on-chain credit and attestation schemes have lingered as concepts for three years is not technical difficulty. It is that no rational person volunteers to carry a permanent, portable, censorship-reactive record of every transaction they have ever touched, and no rail depending on voluntary self-branding will ever be complete. That's a philosophical trap wearing an engineering costume, and it is exactly the trap that sanctions-driven design keeps walking into.
The prevailing read is that designating the ICC hurts the ICC. Backwards.
What a designation actually does is certify, in the most credible venue available, that dollar rails are political instruments rather than neutral plumbing. That argument has been made for years by Moscow, by Beijing, by every sanctions-hammered treasury on earth. It has always been discounted because the messenger had an obvious interest. The Hague does not have an obvious interest. The Hague has 125 member states, a mandate that European capitals do not seriously dispute, and a payroll that needs to clear on time.
Watch what that unlocks. The Assembly of States Parties does not need to denominate contributions in a new currency to route around the problem. It needs one credible non-dollar settlement path for a portion of its operating budget. Between mBridge, the euro system's own cross-border initiatives and the quiet expansion of bilateral local-currency arrangements, that path exists in prototype. A court is the best possible proof of concept, because a court is the least threatening possible user.
And the rail does not check credentials. Whatever gets built so the ICC can pay interpreters in The Hague will not be engineered to exclude the next sanctioned party, or the one after that. You cannot build a channel that only the righteous can use — the design properties that make it resilient make it indiscriminate. The ICC is the wedge, not the destination.
There is a second untold consequence, closer to home for anyone in this industry. If Brussels is forced to choose between enforcing its Blocking Statute against Dutch banks and preserving the dollar access those banks depend on, the answer will be decided by balance sheets rather than by law. The Blocking Statute has never been tested at this scale. It may not survive contact. And the precedent set in Amsterdam will govern how every EU-domiciled exchange, custodian and payment processor handles a US designation for the next decade.
Watch whether any ICC-adjacent address or entity lands on a designation list, because that is the moment this stops being a diplomatic story and becomes an on-chain compliance event with a traceable footprint. Then watch what the Assembly of States Parties does to its payment infrastructure across the next two funding cycles — a shift of even twenty percent of a €200 million budget off correspondent rails is a signal with a long tail. And watch the Dutch banks most closely of all, since their public statements will arrive long after their private decisions.
The court will survive the designations. Institutions generally do. The interesting question is what gets built in the gap between what the sanctions legally require and what compliance officers decide they can afford to touch — and whether, five years from now, anyone will be able to tell the difference.