Transparency for Borrowers: The Register the United States Chose to Delete

The International Monetary Fund (IMF) completed a guidance note on 25 June 2026, released the following month, instructing its staff on how to build financial crime into surveillance, financial sector assessments and lending decisions. Seven weeks later, on 11 August, the Financial Crimes Enforcement Network (FinCEN) published a final rule removing the obligation on US companies to report who owns them. Most of what had already been reported will be erased. The two documents address different audiences and neither mentions the other. Yet read together they describe a transparency regime whose force is calibrated to a country's need for external money.

That is a harder proposition than the familiar complaint about American backsliding. The question is not whether the United States has diverged from the international standard. It is what the standard turns out to be made of once a member with no need of the Fund declines to follow it. That is a question of power, not of law.

The upgrade rested on two mechanisms

Recommendation 24 of the Financial Action Task Force (FATF) has never required a public register. It requires countries to ensure that competent authorities can obtain adequate, accurate and current beneficial ownership information, and it contemplates a combination of sources rather than a single repository. That flexibility is the reason the American position held together for as long as it did.

The 2016 mutual evaluation rated the United States Non Compliant on Recommendation 24, finding that company formation practice left ownership information unlikely to be adequate, accurate or timely. On 26 March 2024 the FATF published its seventh enhanced follow-up report and upgraded the rating to Largely Compliant, leaving the country with nine Recommendations rated Compliant, twenty-three Largely Compliant, five Partially Compliant and three Non Compliant. The upgrade was credited to the Corporate Transparency Act reporting rule, which operates alongside the 2016 customer due diligence rule obliging covered institutions to obtain ownership information from legal entity customers. The rating rested on the two together; neither would have carried it alone.

Both mechanisms have now narrowed

The first change came quietly. On 13 February 2026 FinCEN issued order FIN-2026-R001, granting covered institutions relief from the requirement to identify and verify the beneficial owners of a legal entity customer at every new account opening. Identification is now required when an entity first opens an account, when the institution learns something that calls the reliability of held information into question, and where its own risk-based procedures demand it. This is not abolition. It is a change of cadence, and cadence is what determines whether held information is current.

The second change is structural. The final rule was published at 91 Federal Register 52508 and took effect on 14 August. It exempts all domestically formed entities from reporting; exempts reporting companies from submitting information on US person beneficial owners and company applicants; and releases US persons from any obligation to update the information behind a FinCEN identifier. The Secretary of the Treasury exercised the exemption power at section 5336(a)(11)(B)(xxiv) of title 31, which requires written concurrence from the Attorney General and the Secretary of Homeland Security. That concurrence was obtained. Whatever else may be said of the rule, it was not procedurally casual.

Deletion is the part with no precedent. FinCEN intends to work with the National Archives and Records Administration to remove records associated with any individual it reasonably believes to be a US person, identifying them by documents such as a passport or driving licence. Filers need take no action, will receive no confirmation, and the exercise is to be conducted once rather than periodically; information filed after 10 February 2027 will not be removed. A reporting population the agency once put at more than thirty-two million entities is now an estimated twenty-eight thousand foreign companies. The record is not being frozen. It is being unmade.

The surviving mechanism is also the weakest. By the end of 2025 FinCEN had received around thirteen thousand reports from foreign reporting companies, and the preamble records its estimate that some fifteen thousand existing foreign companies had still to file. The prong the United States has retained is the one it has never made work.

Registries are what borrowers build

Set that against what the Fund asks of the countries it lends to. The guidance note lists, among illustrative types of conditionality, the establishment of a beneficial ownership registry accessible to competent authorities and to reporting entities for due diligence purposes. Among measures aimed at effectiveness it lists operationalising a registry that is accurate, current and open to the public. A coverage threshold of eighty to ninety per cent of entities may be attached.

These are not hypothetical. Mozambique was required to publish a decree law ensuring collection of adequate, accurate and up-to-date ownership information in line with Recommendation 24. Côte d'Ivoire was required to operationalise a centralised register, adopt internal verification procedures, cross-check entries against tax authority data and give competent authorities direct electronic access. Panama was required, as a prior action, to compile information held by law firms acting as resident agents and upload at least eighteen thousand companies. Seychelles was asked to widen access to its central database; São Tomé and Príncipe to publish the ownership of procurement contract winners.

Each of these is a condition of money. Not one of these countries was being asked for what the United States had until 14 August. Each was being asked for considerably more: verification, reconciliation against tax records, public accessibility, and a measurable share of the corporate population covered. The American register never carried a coverage threshold, never verified what it received, and has now been withdrawn and set for deletion. Less was asked, and even that has gone.

There is an obvious objection. Conditionality is a lending instrument rather than a standard, and lenders have always attached terms; the standard-setter is the FATF, which retains the power to list a member that falls short. That objection would carry more weight if the power had ever been used against a major economy. It would carry more weight still if the two processes were independent.

They are not. The Fund's guidance directs staff to treat mutual evaluation reports as the principal evidence of effectiveness, so the peer review supplies the finding and the loan supplies the remedy.

The Fund cannot compel Washington

The guidance note is candid about the limits of its own reach, and the candour repays reading. On multilateral surveillance it states that the Fund "may not require a member to change its policies" in the interests of the international monetary system; it may discuss the effects of those policies and suggest alternatives. Conditionality, the instrument with genuine force, applies only where a country is drawing on Fund resources. Financial sector assessments are scoped to financial stability rather than compliance. Article IV coverage becomes mandatory only where financial integrity issues are macro-critical.

The note then anticipates precisely the situation now in view. It observes that in some advanced economies remaining gaps, expressly including entity transparency and enforcement against foreign proceeds of crime, may not reach the threshold of macro-criticality. Engagement then proceeds voluntarily. The same passage acknowledges that such gaps could nonetheless produce spillovers for other jurisdictions. The Fund has written down the reason it will not be able to press the point, and has done so in a document intended to strengthen its engagement. The limit is stated, not discovered.

The rest of the note makes the omission conspicuous. It states that advanced economies are frequently destination jurisdictions for laundered funds. Where assets are not detected, confiscated and recovered, the safe-haven effect entrenches illicit economies and weakens state capacity in the countries the money came from. It also records that staff assessment of vulnerabilities draws principally on FATF mutual evaluation reports. That places the Fund's view of the United States downstream of a peer review the United States helped design.

Treasury has priced only one side

The most durable feature of the American position is that it does not deny the value of the data. The preamble to the final rule acknowledges compelling reasons for broad ownership reporting and notes that Treasury has advanced them itself. Its case is narrower and more durable: the Corporate Transparency Act contains a statutory direction to minimise burden, and that direction now outweighs the rest. Requiring reports from domestic entities, Treasury concluded, "would not serve the public interest".

That case arrives with figures attached. Treasury puts the saving to entities formerly classified as domestic reporting companies at around eighteen billion dollars since the interim rule. It estimates an annual reduction of roughly fifty-three million burden hours and a yearly cost reduction of about nine billion dollars. The preamble engages directly with commenters who invoked Recommendation 24 and warned of grey-listing. International standards are treated as one input into a balancing exercise, not as a constraint upon it.

Nothing on the other side of that ledger has been costed with comparable precision. The IMF note asserts that money laundering can undermine stability, growth and institutional quality, and leaves the assessment of severity to staff judgement supported by indicators. The quantitative literature it cites measures the effect of grey-listing on capital flows, which is the price of being named rather than the price of the underlying gap. An argument expressed in dollars will beat an argument expressed in principle, and it will keep beating it until somebody does the arithmetic.

Firms inherit an unpriced obligation

The practical consequence runs through jurisdictional risk assessment rather than through file-level due diligence. Independent ownership data is becoming available in inverse relation to a jurisdiction's economic weight, because the mechanism that produces it is loan conditionality. Firms whose country risk models treat register quality as a proxy for jurisdictional risk will therefore score low-income borrowers favourably and major financial centres poorly. That is the opposite of what those models exist to capture. It is also the opposite of the exposure the Fund itself describes.

Timing sharpens this. The FATF confirmed in its 2024 follow-up report that the United States would next report on progress through its fifth round mutual evaluation. The FACT Coalition states that the evaluation is under way, with publication expected later this year. A firm that waits for that report before adjusting its treatment of American entity structures will be reacting to an assessment of arrangements dismantled while the assessors were working. The findings will be accurate and already historic.

Transparency has become a loan condition

The register has not disappeared from the international system. It has migrated. Countries that need money build registers with verification requirements, cross-checks against tax data, public access and coverage thresholds written into their programmes; the largest shareholder in the institution attaching those conditions has closed its own and arranged for the contents to be destroyed. Both outcomes flow from the same body of standards, applied by institutions with the same nominal membership.

A rule that binds in proportion to indebtedness is not a standard. It is a covenant, and covenants are enforced by creditors against debtors and by nobody in the other direction. Compliance functions have spent a decade treating beneficial ownership transparency as a technical problem awaiting better data. The events of this summer suggest it was never a data problem but a question of who could be made to answer. That question has never reached the jurisdictions where the money finally comes to rest.

Do you know whether your jurisdictional risk model rewards a country for having a register, or for being poor enough to have been made to build one?

If that distinction has never been tested, we can help you establish the position before an examiner does; contact us.

At OpusDatum we test whether entity risk assessments reflect where ownership information can actually be obtained and relied upon, rather than where a framework says it should exist. Our work examines the evidence a supervisor expects when a firm explains why a structure formed in a major financial centre was treated as lower risk than one formed in a jurisdiction under a Fund programme.