25 June 2025 · Yair Knijn

The PEP you decided wasn't worth flagging just took office

The compliance officer who cleared this case did nothing careless at bind. The party was a private businessman, no public office, no government role. The screen returned a possible PEP match against a name floated for a ministry years earlier, the officer checked, confirmed the appointment never happened, and dispositioned it not a PEP. Clean call, documented, closed.

Two years later the same individual was appointed to a state procurement board. Nobody re-ran the check, because in this shop PEP status was something you decided once, at onboarding, and stored as an attribute on the record. The policy was still live. The flag never came back.

PEP exposure is a status that turns on and off, not a permanent label

The mistake is treating "PEP" like date of birth, something true or false for the life of the file. It isn't. It is a state tied to a role, and roles change. A private party gets appointed. A sitting official resigns or loses an election. A family member marries into a political dynasty. Each of those is a transition, and the transition is exactly when your screen needs to fire.

Suppress the flag at bind because "they aren't a PEP today" and you encode today's answer as if it were permanent. The relationship outlives the answer. The day the answer flips, your file is still asserting the old one.

FATF's expectation: enhanced due diligence and ongoing monitoring

FATF Recommendations 12 and 22 are explicit that PEP measures are not a one-time onboarding gate. They require enhanced due diligence for higher-risk PEPs, senior sign-off, source-of-wealth work, and ongoing monitoring of the relationship, not a single check at the start. The framework also expects someone who leaves office to be treated as a PEP for a risk-based period, at least 12 months and often 12 to 24 under national rules, because exposure does not vanish on the last day in post.

The UK FCA's June 2025 finalised guidance on PEPs (FG25-3) pushes the same way: status and risk get reviewed proportionately over the life of the relationship, not frozen. The standard cuts both directions. You re-evaluate when a party becomes exposed, and step measures down, on evidence, when they cease to be.

Why insurers under-weight PEP risk and what that costs at claim time

Insurance treats screening as a binding-day formality more than banking does. No daily transaction stream, the premium clears once, and the policy sits quietly for years. That quiet is the problem. A life policy or a large surety bond can run the full window in which a party's political exposure appears, matures, and becomes the reason the money is moving.

The bill arrives at claim or surrender. That is when funds leave, when source-of-wealth questions get sharp, and when your MLRO asks why a now-obvious PEP was carried for two years with no enhanced measures. "We screened them at onboarding and they weren't a PEP then" is not the defence it sounds like. It documents the exact gap: you knew the rule was continuous and you ran it once.

Re-evaluating PEP status on every re-screen and surfacing the change

The fix is to stop storing PEP as a settled attribute and treat every re-screen as a fresh adjudication. Concretely:

• Re-derive PEP status on each periodic and event-driven screen. Never inherit the onboarding disposition as fact.
• Diff against the prior result so a not a PEP that becomes a PEP surfaces as a status change, not buried in a fresh match list.
• Route the transition to enhanced due diligence: refreshed source of wealth, senior sign-off, a documented decision to continue or exit.
• Apply the leaving-office clock deliberately, running measures through the risk-based tail instead of dropping them when a term ends.

InsureGuardAI re-adjudicates PEP status on every re-screen instead of trusting the bind-day call, and a workspace surfaces the transition with the prior disposition next to the new one, so the moment your cleared party takes office lands as an alert in the case file rather than a surprise at claim. That is the difference between screening once and screening for the life of the policy.