PEP Screening

A politically exposed person holds, or has held, a prominent public function. Political exposure alone does not establish criminal conduct by the identified person. FATF guidance on politically exposed persons treats the additional measures as preventive safeguards against financial crime. A screening match therefore requires examination before an institution decides what the finding means for its customer.

The screening population extends beyond the person signing an account application. Corporate customer verification can identify beneficial owners whose public roles require examination. Family relationships and close business associations need evidence that fits the applicable legal definitions.

The AMLR generally applies from 10 July 2027, with specific measures for politically exposed persons. Article 42 addresses senior management approval, the source of wealth and funds, and more intensive ongoing monitoring. Enhanced due diligence connects those findings with the controls required for the relationship.

AMLA submitted final draft customer due diligence standards to the European Commission on 1 October 2026. The drafts include PEP screening, family members and close associates. Final drafts still require adoption before they become binding delegated legislation.

Entity resolution links records that may refer to the same person across different data sources. A screening implementation can compare name variants alongside birth information and verified identity attributes. The reviewer needs to see which attributes support a match and which contradict that proposed identity.

Machine learning can help rank candidate matches for review when datasets contain spelling variations or incomplete records. The institution should test those rankings against representative names, languages and missing-data cases. Identity orchestration can carry the verified attributes and review status into the customer onboarding workflow.

Names written in different scripts can produce several plausible spellings in a screening dataset. Wolfsberg guidance on multilingual screening distinguishes transliteration from translation and explains the resulting matching difficulties. A useful review preserves the original spelling alongside the transformed version used by the matching system.

A score alone does not explain why two records describe the same person. Reviewers should compare the supporting attributes and record the reason for confirming or dismissing the candidate. Model validation can examine the matching behavior and the effects of changed thresholds before deployment.

Graph-based analysis can connect people, public roles and companies within a documented relationship model. The model can help a reviewer trace a beneficial owner through several company records. Each connection should retain the underlying source and the period during which the relationship was evidenced.

A shared photograph or an unverified online mention does not establish a legally relevant close association. Teams should distinguish a recorded corporate relationship from a suspected family connection before assigning screening significance. A graph presents the evidence available to the reviewer without deciding the legal classification automatically.

Political exposure can change after an institution accepts a customer or identifies a beneficial owner. Perpetual KYC can route a new public appointment or changed ownership record into a further customer review. The workflow needs a clear record of the new information and the decision that followed.

Institutions should examine the freshness of external data alongside the reliability of the original source. Leaving a public role does not automatically resolve every risk associated with the customer relationship. A review should assess the remaining exposure under the applicable rules and document the controls retained.

PEP screening examines political exposure, while sanctions screening examines restrictions attached to designated persons or other relevant circumstances. Both processes may use similar matching technology, but the legal questions and resulting actions differ. An institution should retain separate reasons for each finding when a customer appears in both workflows.

Technology buyers should test the full path from an alert to a documented decision. Useful cases include similar names, incomplete birth information and a corrected external record. The test should establish whether the system preserves the evidence, review history and current customer status.