A PEP or RCA match needs evidence before a decision

A PEP result points to public-function exposure; an RCA result points to a relative or close associate in the data. Neither label confirms that the matched record belongs to your customer. Neither is an allegation of wrongdoing. Review identity first, then the classification and any relationship evidence, and document what remains unresolved.

This guide offers a way to read a screening result, not a universal legal test. The applicable rules and your institution’s procedures determine the measures required for a particular relationship. A useful case file makes those decisions traceable instead of hiding them behind a single coloured status.

Begin with the person. Compare the full name and available aliases against the information you hold. Then examine date of birth, nationality, location and other available identifiers. A similar name is a candidate; it is not a completed identity check. Equally, a missing date of birth does not establish that the candidate is someone else.

Next, ask why the record is classified. For a public office-holder, look for the function, institution, country and available dates. For a relative or associate, look for the relevant person and the documented relationship. Keep a note of whether you saw the underlying source or only a summary of it.

Finally, ask what the evidence supports. A page showing a public appointment supports a claim about that appointment. It does not, by itself, establish every family relationship or prove the identity of your customer. A well-written note separates these conclusions so another reviewer can follow the reasoning without repeating the entire search.

The distinction matters when a result looks persuasive. A familiar surname, a photograph and a recognisable institution can encourage a quick conclusion. None removes the need to compare identifiers. Write the unresolved question explicitly: for example, “The source names the same person, but no birth date is available for comparison.” That is a useful finding, even when it does not close the case.

FATF’s PEP guidance describes preventive controls and warns against equating PEP status with criminal activity. It also explains that a commercial database alone does not satisfy the relevant requirements. Treat the database as an input to your review.

A sanctions designation needs its own assessment against the relevant list and rules. An RCA label should not silently become a sanctions finding in a case note. Where several categories appear, preserve each category and its supporting source. The reviewer should be able to tell which conclusion came from which evidence.

Consider an illustrative case, not a real customer. Your customer and a database record have the same name. The record is tagged RCA and links to an office-holder as a spouse. The record has no birth date, while your customer file does. You have a relationship claim to inspect, but insufficient identity evidence to conclude that it concerns your customer.

Open the cited material and check that it actually names both people and describes the relationship. Note whether the page is an official biography, a disclosure or another source. If the link no longer resolves, keep that limitation in the case record; a broken link does not confirm or disprove the relationship.

A practical review note for this example could read: “Candidate shares the customer’s name. Source describes a spouse relationship with an office-holder. No comparable birth date was available. Identity remains unresolved; further identifying evidence requested.” This is a suggested note format, not an automated conclusion produced by the platform.

If the underlying material refers to a different person, explain the distinguishing evidence when rejecting the candidate. If it supports the match, document the evidence used to accept it and continue with the applicable risk review. Do not delete the distinction between matching a person and assessing their relationship merely because both steps happen in one screen.

A source may contain an appointment date, an end date and a publication date. Your screening record also has a check time. Those dates serve different purposes. A check performed today can return historical information, and a recently retrieved page can still describe an old appointment.

Copy dates with their meaning. “Office ended in 2022” is different from “page retrieved in 2026.” Where an end date is absent, record that absence rather than supplying a date from memory. Where the source explicitly identifies a former role, preserve that context in the review.

Avoid a universal countdown in a case note unless the applicable policy and rules support it. This article does not prescribe a period after which a former office-holder or associate ceases to require attention. The point is to keep enough dated evidence for the responsible reviewer to apply the correct policy.

ScreenVeritAI presents PEP and RCA findings in one results tab, with separate classifications so that an RCA-only person is not presented as an office-holder. Where the record contains relationship evidence, the reviewer can inspect the available connection, sources and dates. Both labels can appear for the same person.

The data pipeline uses explicit PEP and RCA classifications from the OpenSanctions graph. Its documentation on using PEP data distinguishes the enriched default collection from the narrower office-holder collection. A source classification and a resolved relationship are separate pieces of evidence: not every classified RCA record has a relationship that can be displayed in full.

This is screening of available records. It does not infer a family relationship from matching surnames, discover an entire ownership structure, or establish that every relationship is current. When information is absent, the review still needs to say what is missing.

The verified production snapshot dated 30 September 2026 contains:

The arithmetic is 695,260 + 35,508 − 454 = 730,314. These are counts for a dated snapshot, not a promise of complete worldwide coverage. The 297 datasets are contributing collections; they should not be described as 297 independent public authorities. The coverage page for this screening explains the product scope and its limitations.

Before closing a candidate, preserve the query, check date, record identifiers and available source links in your review process. Add the identifying evidence that supported or ruled out the match. Then distinguish confirmed information from unresolved information and record who made the decision.

For the illustrative spouse case, that means retaining both parts of the assessment: whether the source supports the relationship and whether the named individual is your customer. “RCA match reviewed” alone tells the next reviewer neither. A short explanation of the evidence is more useful than a longer narrative with an unsupported conclusion.

If no candidate appears, describe the result as no match in the sources checked. Retain enough context to understand the search later, including the name variant used. Further checks should follow the risk and evidence in the case, rather than a promise that one search can settle every possible connection.

The next action should address a specific gap: obtain a missing identifier, inspect an unavailable source, clarify the relevant role or refer the case for review under your policy. A screening result is most useful when it makes that next action clear.