What if compliance ran inside settlement?

Compliance and settlement are two different systems in almost every financial institution on earth. One decides whether money may move. The other moves it. They are built by different teams, run on different infrastructure, and meet only at the edges: a screening gate before the payment leaves, an investigation queue after it lands.

This page is about what that separation costs, and about the design question the industry is starting to ask out loud: what would change if the rules ran inside the settlement itself?

How it works today

Follow a cross-border payment and you can watch compliance chase it.

Before release, the originating institution screens the payment: sanctions lists, politically exposed persons, jurisdiction rules, internal limits. Under the FATF’s Travel Rule, originator and beneficiary information must travel with it. Then the payment enters the correspondent chain, and here the chasing multiplies: each intermediary bank runs its own screening on the same payment, against its own lists, under its own risk appetite. Any hit at any hop stops the payment, often for days, usually for reasons invisible to the sender.

After settlement comes the second system: transaction monitoring. Rules and models sweep settled payments for suspicious patterns and raise alerts for human review. The alerts arrive after the money has moved. If something was truly wrong, the remedy is investigation, reporting, and attempted recovery of value that is already gone.

Note what the settlement systems themselves did in this story: nothing. The rails that moved the money, correspondent accounts, clearing systems, ledgers, are blind to every rule involved. They will settle anything that arrives with valid formatting. All enforcement lives in systems bolted on before and after, which is why the enforcement is simultaneously slow and porous.

What the separation costs

The direct bill is documented. LexisNexis Risk Solutions put the global cost of financial crime compliance for financial institutions at more than $206 billion a year in its 2023 global study, with 98% of institutions reporting cost increases. That is the spend on people, screening technology, and process, before a single fine.

What the spend buys is a system that mostly alarms on innocents. Industry analyses consistently find that around 90 to 95% of transaction monitoring alerts are false positives, each one consuming investigator time and, frequently, delaying or blocking a legitimate payment. The FSB’s work on cross-border payments has repeatedly identified compliance processing among the frictions keeping the G20’s speed and cost targets out of reach; in October 2025 it concluded those end-2027 targets are unlikely to be met at the global level.

There is a subtler cost too. Because every intermediary repeats the checking, the checking compounds with the length of the chain. The payment is screened not once but at every hop, serially, by institutions that cannot see each other’s conclusions. Compliance done this way scales with the number of intermediaries, and the correspondent model supplies plenty of them.

The design alternative

The alternative keeps every rule and moves the enforcement to a different place.

Programmable settlement infrastructure makes it possible to attach conditions to the settlement event itself: the transfer executes only if its governing policy is satisfied. Sanctions posture, counterparty limits, jurisdictional rules, permitted assets, required information traveling with the payment, all of it evaluated at the moment of settlement, as a condition of settlement. A transfer that cannot satisfy the policy does not settle. There is nothing to unwind, no window between movement and verification, and no need for each party in a chain to re-check what the settlement itself has proven.

Three properties follow, and they are worth stating precisely.

Enforcement becomes structural. Today a rule is enforced if a screening system happens to catch the payment in time. Inside settlement, an unsatisfied rule means no settlement occurred. The failure mode changes from “money moved that should not have” to “money did not move”, and the second failure is recoverable in a way the first is not.

Evidence becomes a byproduct. When policy is evaluated in settlement, every settled transaction can carry proof that its conditions were met, in the form of a cryptographic attestation a third party can verify without seeing the underlying business data. Audit stops being archaeology.

The compliant majority speeds up. Most payments are innocent. Today they queue behind processes designed for the exceptional few. When the check is the settlement condition rather than a separate stop, a payment that satisfies its policy settles at the speed of the rail, and the exceptions are stopped cleanly at the boundary.

None of this removes the human parts of compliance: risk appetite, list quality, investigation of the genuinely suspicious. It removes the structural gap between the rules and the rails, the gap all that $206 billion is spent patrolling.

Where Frame fits

Frame is the settlement layer for global finance: one integration to orchestrate payments at scale across fiat rails, stablecoins, and tokenized deposits, with compliance enforced on every transaction and settlement in seconds instead of days.

Compliance enforced in settlement is Frame’s founding design decision. Frame’s Rules Engine evaluates every transaction against its governing policies inside the settlement flow, and a transfer that cannot satisfy them does not settle. Every settled transaction produces verifiable evidence that its conditions were met, without exposing the underlying business data. Because Frame is rail-neutral, that enforcement travels with the payment whichever rail carries it: a fiat network, a regulated stablecoin, or a tokenized deposit.

For banks, payment providers, exchanges, platforms, and enterprises, the question this page began with stops being hypothetical. Compliance inside settlement exists as infrastructure; what remains is deciding which policies should govern each payment, and letting the settlement enforce them.