A Wolfsberg correspondent banking AML assessment SAMA exercise tends to fail in one place: the gap between what the risk assessment says and what the screening team actually does. The questionnaire asks for control-level description. An independent assessor then tests whether your documents, procedures and files support each answer.
This article explains how to prepare for that cross-check. It sets out the four customer-screening areas you should reconcile, and a three-column mapping that shows policy governs practice.
A note on precision first. The Wolfsberg Correspondent Banking Due Diligence Questionnaire is structured by section, and section numbering has changed between versions. SAMA's AML/CTF rules are also numbered. Confirm the exact section and article references against the versions your counterparties and your assessor use. The substance below does not depend on the numbering.
Why the misalignment appears
Three patterns recur in financial crime programs:
- The enterprise-wide risk assessment names customer and geographic risk factors, but the screening procedure does not use them to set screening behavior.
- Policy describes screening at onboarding and ongoing, but operational teams run it on different triggers or different lists.
- Questionnaire answers are drafted by one team from policy text, while the procedure owners are never consulted.
None of these is a screening failure by itself. Each becomes a finding when an assessor cannot trace a stated control to a document and then to evidence.
The four customer-screening descriptions to reconcile
The questionnaire's customer-screening area can be read as four control descriptions. For each, your risk assessment must say the same thing as your procedure.
1. Scope of who is screened
State which parties are screened: customers, beneficial owners, controllers, authorized signatories, and where applicable the respondent bank's own ownership. Your risk assessment should explain why that scope is proportionate. Your KYC procedure should show the data fields that feed screening, so each party in scope is actually captured.
Typical gap: policy says beneficial owners are screened, but the onboarding system only passes the legal entity name to the screening tool.
2. Lists and regimes used
Describe which sanctions, PEP and adverse media sources are applied, and how they are maintained. For a Saudi institution this normally includes UN, local and the international regimes your business exposure requires. The risk assessment should justify the choice of regimes, and the procedure should name the owner of list updates.
Typical gap: the questionnaire names regimes that the procedure does not mention, or lists are updated without a recorded owner or date.
3. Timing and frequency
Explain when screening occurs: at onboarding, on list changes, on trigger events and periodically. Cross-check this against the risk-based review cycles in your risk assessment. If higher-risk customers are meant to be reviewed more often, the procedure must show it.
Typical gap: a single batch cycle applies to every customer while policy implies risk-differentiated frequency.
4. Alert handling and escalation
Describe how alerts are reviewed, who can close them, what must be escalated, and how decisions are recorded. The risk assessment should link escalation thresholds to your stated risk appetite. Assessors will sample alerts and look for consistent rationale and sign-off.
Typical gap: closure rationale is free text with no standard, so similar alerts are decided differently.
The three-column mapping
The mapping is a single working document. Each row is one claim you make in the questionnaire or in the risk assessment.
- Column 1, Requirement and claim: the SAMA AML/CTF requirement or Wolfsberg question, and your answer in the words you submitted.
- Column 2, Policy and risk assessment source: the document, section and version where that claim is stated or justified.
- Column 3, Procedure and evidence: the operational procedure, system configuration, owner, and a sample of evidence such as a customer file or alert record.
Build it in this order:
- List every screening statement from the questionnaire draft and the risk assessment.
- Match each to a policy clause. If none exists, record a policy gap.
- Match each policy clause to a procedure step and a named owner. If the step differs from the policy, record a procedure gap.
- Pull a small set of real records for each row to confirm the procedure operates as written.
- Resolve each gap by changing the document, the practice or the questionnaire answer, never by leaving it unrecorded.
A row with all three columns filled and evidence attached is defensible. A row with an empty third column is what an assessor will find first.
Reconciling with SAMA expectations
SAMA's AML/CTF rules expect a documented, risk-based approach in which the institution's risk assessment drives its controls. The mapping makes that link visible. For each screening control, the question to answer is: which risk identified in the enterprise-wide assessment does this control address, and where is that stated?
Where your Wolfsberg answer is more detailed than your SAMA-facing documentation, align the documentation. Where it is more optimistic than your practice, correct the answer.
What an independent assessor will do
Expect the assessor to read the risk assessment and policies first, then test customer files, alerts and reports against them. That means your mapping should be finished before fieldwork, not assembled in response to a request. Keep the evidence samples traceable, and make sure the people named as owners know their procedures will be read.
How we can help
Our AML Assessment is an independent assessment of your AML/CTF program against the SAMA rules and the FATF Recommendations. It also draws on Wolfsberg as a reference framework. The work follows four stages: plan, review, test and report. The review stage covers policies, risk assessment and governance, and the test stage covers customer files, alerts and reports.
The assessment includes an enterprise-wide ML/TF risk assessment review, policy and procedure review, and sample testing of customer files and alerts. It is designed to serve as independent testing of your AML/CTF program. The sample is risk-based and agreed with you, typically covering every high-risk segment.
You receive an independent AML/CTF assessment report, findings rated by severity, a remediation roadmap and a board summary. The assessment takes 5–8 weeks.
If your next Wolfsberg response or independent assessment is approaching, contact us to discuss scoping an AML Assessment, so that the policy-to-procedure mapping is tested before someone else tests it.