When card fraud is detected, the quality of the investigation is largely decided in the first hours. Logs roll over, terminal configurations change, and people who touched the data go off shift. This article sets out a practical approach to SAMA CFF forensic evidence for a card fraud investigation in Saudi Arabia: six artifacts to secure early, and a single chain-of-custody record that serves both internal audit and law enforcement.
A note on scope. The SAMA Counter-Fraud Framework (CFF) expects regulated institutions to investigate fraud incidents and to keep the records that support that work. We refer to the incident-investigation requirements in general terms here rather than by article number; confirm the exact wording against your institution's current copy of the framework and your legal counsel's guidance. The four-hour window is a working target for your team, not a figure quoted from the framework.
Why the first four hours matter
Most evidence in a card-fraud case is volatile or controlled by someone else: switch logs held by a processor, terminal data held by an acquirer, device data held by a fraud platform with short retention. If you wait for a full investigation plan, some of it will be gone or altered. If you collect it informally, a defence lawyer or an auditor can fairly ask whether it is the same data that existed at the time.
Two failures recur:
- Incomplete collection, where the team captures transactions but not the context that explains how they were authorised.
- Broken custody, where nobody can state who exported a file, from which system, with what query, and where it has been since.
The six artifacts
1. Transaction authorization logs with timestamps
Export the authorization records for the affected cards and the suspect time window, including approval and decline codes, amounts, merchant identifiers and the timestamps from each system. Record the time source and time zone of every system. Mismatched clocks are the most common reason timelines cannot be reconciled later.
2. Cardholder authentication attempts and failures
Capture the full sequence of authentication events: one-time passcode requests, PIN failures, 3-D Secure challenges, enrolment changes and any contact-detail updates shortly before the fraud. The order of failures and successes often shows whether the fraudster held credentials or was guessing.
3. Geolocation and device-fingerprint data
Preserve IP addresses, derived location, device identifiers, app or browser attributes and session records linked to the transactions and to the authentication events. Note the retention period of the source platform so you know how urgent the export is.
4. Merchant-terminal configuration snapshots
For card-present cases, take a snapshot of the terminal configuration, software version, key-injection or parameter records and recent change history for the terminals involved. Where the terminal belongs to a third party, send a written preservation request the same day and log it.
5. Issuer-switch message flows
Obtain the message-level records between the acquirer, the switch and your issuing systems for the disputed transactions. These show what was actually sent and approved, including response codes and any stand-in processing. Request them from the switch operator in writing if you do not hold them directly.
6. Fraud-rule triggering events
Preserve the rule engine's alerts, scores and the rule versions in force at the time, along with any analyst actions on those alerts. This evidence shows what your controls saw, when, and what was done, and it is the first thing a regulator will ask about.
One chain-of-custody record, not two
Internal audit wants to see control. Law enforcement wants to see integrity. Both are satisfied by the same record if it is designed once. Use a single log per case, with one line per item collected, containing:
- Case reference and artifact number (1 to 6)
- Source system, owner and the exact query or export method used
- Collector's name, role and the date and time of collection, with time zone
- Hash value of the file at the point of collection, and the tool used to compute it
- Storage location and access restrictions
- Each transfer or access event: who, when, why, and the hash re-checked
- Reviewer sign-off and, where applicable, the handover receipt to the authority
Keep the working copy separate from the preserved original, and analyse only the copy. Store the log in a system with restricted write access and its own audit trail, so that the record itself cannot be edited without trace. Because the audit view and the law-enforcement view are simply filters on the same log, nobody maintains a parallel system.
A four-hour sequence
- First hour: declare the incident, appoint a single evidence lead, freeze log deletion and rollover on the relevant systems, and open the custody log.
- Hours one to two: collect artifacts 1, 2 and 6, which usually sit in systems you control.
- Hours two to three: collect artifact 3 and send written preservation requests for artifacts 4 and 5 to any third party.
- Hours three to four: hash and secure everything, record gaps and pending requests honestly, and brief the fraud and compliance leads on what has been preserved.
Gaps are acceptable if documented. Silent gaps are not.
Where specialist support fits
Fraud teams are strong at pattern analysis and weaker at forensic handling under pressure. Our Digital Forensics & Incident Investigation service is built for that moment: our specialists help contain an incident, preserve evidence, investigate root cause and scope, and support regulatory reporting and recovery. Our process runs in four stages: triage, preserve, analyze and report. The preserve stage covers evidence collection with chain of custody.
The service can be arranged ahead of time. A pre-agreed incident response retainer means we can start immediately, with contracts and access already in place. Response is on demand, measured in hours. Deliverables include an investigation report, an incident timeline, indicators of compromise and remediation recommendations.
To make the first hours smoother, our Incident Response Readiness & Tabletop Exercises service prepares your organization to respond, including roles, escalation and regulatory notification, so that everyone knows who does what in the first hours of an incident.
Next step
Review your last card-fraud case against the six artifacts and the custody fields above. Note which items you could not have produced and why. That list is the starting point for a retainer conversation, a playbook update, or both. Contact CISO Consulting to discuss Digital Forensics & Incident Investigation for your institution.
