A customer marked "verified" three years ago was verified three years ago. That is all the record actually proves.
Onboarding checks record what was true on one day. Since then the ID on file may have expired, the customer's details may have changed, and their risk profile may look quite different. Verification is a date, not a permanent status, and reKYC is the discipline of treating it that way without making every customer start from scratch.
Different customers, different review rhythms
Scheduled and triggered
reKYC arrives by two routes. A scheduled review comes due on a date set by the customer's risk tier. A triggered review happens because something changed, whenever it changes.
| Trigger | Example |
|---|---|
| Document expiry | The Resident ID or National ID on file passes its expiry date |
| Material change | The customer updates their name, nationality or identity document |
| Risk change | A transaction pattern or a screening result raises the customer's risk rating |
| Dormancy | An account is reactivated after a long period without activity |
| Periodic | The scheduled review date for the customer's risk tier arrives |
Document expiry is the most common trigger in practice, and the easiest to automate, because the date was captured at onboarding.
Risk-based, not uniform
Reviewing every customer on the same cycle wastes effort on the low-risk majority and gives too little attention to the few who need it. A risk-based approach reviews higher-risk customers more often and lets low-risk customers go longer between checks.
The frequencies themselves are set by the institution and its regulator, not by a vendor default. We deliberately do not state them here; confirm them with your own regulator, the same way retention periods for ID records should be confirmed rather than inferred.
Make it cheap for the customer
reKYC fails when it feels like onboarding all over again. The customer should re-verify through the same short journey they already know — a fresh ID capture and selfie — and be asked only for what is actually stale. If only the ID has expired, a new ID and a live face confirm the same person holds it; the rest of the file does not need to be rebuilt.
That re-verification is a new face comparison, so it is biometric processing under the PDPL each time, exactly as at onboarding.
What the record should hold
For reKYC to work, each verification needs to be recorded as an event, not a flag: when it happened, against which document, which checks ran and with what result, and when the next review is due. With that, the schedule computes itself and an auditor can see why any customer was or was not re-verified.
ekyciq includes AI-driven reKYC triggers and risk-based periodic re-verification, using the same capture as the onboarding journey. See how ekyciq works, or talk to us about your existing customer book.
Related posts
-
10 Automated Checks Every Cheque Platform Should Perform
A single successful OCR result is not enough. Confidence has to be built across the whole document.
31 August 2026 -
Detecting Cheque Alterations and Fraud Indicators with AI
An unusual image characteristic is not proof of fraud. The system flags; authorised people decide.
31 August 2026 -
Why Cheque Processing Needs More Than OCR
Reading a cheque and validating a cheque are two different problems. OCR only solves the first.
31 August 2026


