The audit-ready screening decision record: a runbook for compliance teams

An audit rarely begins with the easy case. It begins with a decision someone made months ago, after the team changed systems, the analyst left, the source was updated and the original browser session disappeared. The question is no longer “can we screen this customer now?” It is “what did we know then, what did we do with it, and can we demonstrate that without rebuilding the entire case?”

That is the job of a screening decision record.

It is not a screenshot. It is not a vendor’s green tick. It is a compact, chronological object that binds together the input, source state, returned candidates, reasoning, human decision and later changes. Once teams use it consistently, audit readiness stops being a recovery project and becomes a property of daily work.

Short answer

An audit-ready decision record retains six things: the subject and query submitted, the source/data state at the time, the candidates and matching rationale, the evidence used to assess them, the disposition and approver, and the complete event history—including later re-screening. The record must preserve uncertainty: a missing identifier, stale source or unverified ownership relationship is part of the evidence, not an omission to hide.

The six parts of a durable record

These six parts should be linked by stable identifiers, not a loose stack of PDFs. A new case should reference the prior case when appropriate without altering the original record.

Start with the question the record must answer

Every field should help a future reviewer answer a real question:

• Which subject did we believe we were screening? Preserve original and normalised input plus identifiers.
• What data did the control rely on at that moment? Preserve sources, state and meaningful limitations.
• Why did the system surface this candidate? Preserve the matching explanation and source record.
• Why did the team decide this outcome? Preserve the reviewed attributes, reviewer and reason.
• What changed later? Preserve rescreening events and do not rewrite history.

If a field cannot answer one of those questions, it may be telemetry rather than compliance evidence. Keep it if engineering needs it, but do not mistake it for the record itself.

Make uncertainty explicit

The most credible decision records do not pretend a control knows more than it does. They can say:

• “No unique identifier was supplied; the disposition is based on name, country and customer-provided document.”
• “Relationship data was retrieved on this date and has a coverage limitation.”
• “A source was stale or unavailable; the policy routed the case for escalation.”
• “The candidate was cleared because documented identifiers conflicted; monitoring will reopen on a material change.”

This is not weak wording. It is the operational truth that lets an approver, auditor or later analyst understand the boundary of the conclusion.

Treat rescreening as a new event, not an edit

Continuous monitoring is often described as a service that sends alerts. The more useful mental model is a versioned decision history. A new sanctions record, source update, ownership change or customer-provided identifier creates a new event. The new review can draw on the prior evidence, but it should produce its own timestamped disposition.

This gives the team two valuable capabilities: it can show why an earlier decision was reasonable when made, and it can show that the decision was revisited when the facts changed. Both are needed for a defensible control.

The monthly quality check

Sample a small number of completed cases each month and ask an independent reviewer to reconstruct them using only the exported decision record. Score:

1. identity clarity;
2. source/freshness clarity;
3. candidate explanation;
4. disposition rationale;
5. reviewer/approval trace;
6. ability to understand later changes.

The gaps tell you whether the problem is data collection, workflow design, training or product integration. This is a much better control-health signal than the number of searches completed.

Where Verifex fits

Verifex’s Evidence Capsule, audit trail and screening-evidence vault are intended to keep the decision record attached to the workflow rather than scattered across tools. Validate exact retention, export and integration behaviour in the current product documentation. The buyer-facing promise should be specific: Verifex helps a team retrieve and explain its own decision record; it does not replace the team’s legal judgment or policy ownership.

Sources and further reading

• Verifex: audit trail
• Verifex: sanctions evidence vault
• Verifex: Evidence Capsule
• FATF Recommendations