Insight · May 2025
The Travel Rule, on-chain: what supervisors actually expect
FATF Recommendation 16 was written for wire transfers and retrofitted onto digital assets. Here is what the rule requires, where on-chain transfers break its assumptions, and what an examiner will ask you to show.
A customer instructs a $40,000 transfer in a tokenized dollar from your institution to an account at another bank. On the legacy side this is unremarkable. The originating bank attaches name, account number, and address to a message, the beneficiary bank receives it, both retain records, and the payment settles. On-chain, the value moves in a single transaction confirmed in seconds, while the information about who is paying whom moves in a separate channel the blockchain itself knows nothing about. Two institutions that may never have spoken now have to reconcile two streams. That gap between the asset hop and the data hop is the compliance problem the Travel Rule poses for digital assets, and it is where examination findings cluster.
FATF Recommendation 16, the "Travel Rule," predates crypto by years. It was written for wire transfers: require originator and beneficiary information to travel with the payment so that law enforcement, sanctions screening, and AML monitoring have something to act on. In 2019 FATF extended Recommendation 16 to virtual assets and virtual asset service providers (VASPs). In doing so it imported a set of assumptions that public blockchains do not natively supply: an ordered messaging network, identifiable intermediaries, and a shared standard for the data. The obligation is old. The operating environment is not.
What the rule actually requires
Strip away the protocol debate and the substance is concrete. When a transfer crosses between two regulated institutions above the applicable threshold, the originating institution must obtain, hold, and transmit a defined set of data to the beneficiary institution, and the beneficiary institution must receive and screen it before or at the point of making funds available.
For the originator, the required fields are name; the account number used to process the transaction, which on-chain is the wallet address; and either physical address, national identity number, or date and place of birth. For the beneficiary, the fields are name and account number, again the receiving wallet address. The originating side has verified its customer. The beneficiary side is expected to have done the same and to receive the counterparty data in a usable form.
Thresholds matter, and they are not uniform. FATF sets the virtual-asset transfer threshold at roughly USD/EUR 1,000. The US implementation through FinCEN under the Bank Secrecy Act applies the funds-transfer recordkeeping and travel rules at $3,000. Where the two diverge, the institution operates to the stricter standard for the corridor it is in, and an examiner will expect to see that you have mapped thresholds to your actual flows rather than applied a single global number for convenience.
Below the threshold, the obligation changes shape rather than disappearing. FATF still expects originator and beneficiary names and account identifiers to accompany the transfer, with verification triggered by suspicion. The small transfer is not a free pass.
Where on-chain breaks the assumptions
Three structural features of public blockchains create the friction supervisors are now probing.
There is no native messaging layer. A traditional wire carries its compliance payload inside the same message that moves the money. A blockchain transaction carries value and almost nothing about the legal persons behind it. The Travel Rule data therefore has to travel on a parallel rail, an interoperability protocol agreed between institutions. Several competing standards exist, and a transfer works end to end only if both sides speak the same one. Coverage is a real operational metric: what fraction of your outbound transfers can actually be matched to a counterparty institution that can receive the data?
Then there is counterparty identification, sometimes called the VASP-discovery problem. Before sending, the originating institution has to determine whether the destination wallet belongs to another regulated institution, and which one, or to a self-hosted wallet, or to an unhosted address of unknown control. A wallet address is not self-describing. Getting this wrong in either direction, treating an obliged institution as unhosted or assuming a regulated counterparty where none exists, produces exactly the control gaps examiners look for.
Then there are self-hosted, or unhosted, wallets. When a customer sends to or receives from a wallet they themselves control, there is no second institution to exchange data with. FATF's guidance does not prohibit these transfers, but it expects the institution to collect the required originator and beneficiary information, apply risk-based controls, and increase scrutiny as exposure grows. This is the corridor most likely to be tested, because it is where the symmetry of VASP-to-VASP messaging does not exist.
What examiners actually look at
Travel Rule compliance is no longer a niche crypto topic. Under ordinary third-party risk management and safety-and-soundness review, it sits inside the same supervisory frame as any other funds-transfer control. Examiners want evidence, not explanations, and the evidence falls into a few predictable categories.
They look for a written, risk-based program: documented thresholds by jurisdiction and corridor, your VASP-identification methodology, and your treatment of self-hosted wallets, all consistent with the FFIEC BSA/AML manual's expectations for funds-transfer recordkeeping.
They look for data quality and transmission proof: that required fields are complete and accurate, that they were actually sent to the beneficiary institution, and that you can produce the message for a specific transaction on request. A policy stating that data is transmitted is worth little without per-transaction artifacts.
They look for the interaction with sanctions and AML. Wallet addresses, originator and beneficiary, should be screened against OFAC's SDN list. Screening should go beyond name-matching to transaction-graph analysis of direct and indirect exposure, including mixer and Tornado Cash exposure, where the code itself can be sanctioned. Travel Rule data should feed your KYT monitoring and SAR decisioning rather than sitting in a parallel silo. Integration with Chainalysis, Elliptic, or TRM Labs is assumed, not novel.
They look for handling of incoming transfers with missing or deficient data: what you do when a counterparty sends without the required fields. A defensible answer specifies a procedure, request, escalate, restrict, or reject, not a shrug.
And they look for records: immutable, exportable, examiner-ready audit trails tying each transfer to its compliance data, retained under BSA recordkeeping rules and producible without a forensic project.
The interoperability problem underneath
The recurring theme is that the Travel Rule is, at bottom, an interoperability requirement dressed as a compliance one. The data only travels if two institutions can connect: same protocol, mutually identified, able to reconcile a value transfer with a separate data message. As tokenized deposits and other on-chain money proliferate inside the regulated perimeter, the industry is building capable but largely disconnected islands of money, and the seam between them is precisely where the Travel Rule lives.
There is no single moment when an institution becomes "Travel Rule compliant." There is the quieter, harder work of building controls that keep digital-asset transfers inside the supervisory framework you already operate under, so that the answer to an examiner's question is a record you can pull rather than a position you have to argue. That is the standard worth designing toward: infrastructure that connects the networks you choose to use without asking you to give up the perimeter, the keys, or the audit trail that make the activity defensible in the first place. That connective layer is the problem Suave is building to address, and we are early enough to say plainly that the work is ongoing.