# Template: security control checklist

One row per control area per system holding personal data. The checklist is the baseline; the evidence column is what makes it real. Copy it and fill it in.

| Control area | The standard to meet | System | Status | Evidence | Last evidence | Owner |
| --- | --- | --- | --- | --- | --- | --- |
| Encryption | At rest and in transit; keys managed and rotated | | | | | |
| Access | Role-based, least privilege, segregation of duties, MFA for privileged access, no shared admin accounts | | | | | |
| Joiner-mover-leaver | Access granted, changed and removed with the role, on a clock | | | | | |
| Logging | Tamper-evident logs of access and change, retained, reviewed by a named person | | | | | |
| Backups | Tested restores, covered by the breach-investigation duty | | | | | |
| VAPT | Before go-live and at intervals; findings with owners, due dates, closure clock | | | | | |
| Sectoral overlays | The frameworks your regulators run, mapped onto the same baseline | | | | | |

Column guide:

- **System**: one set of rows per system holding personal data, including vendor-run systems.
- **Status**: Met, Partially met, Not met. Partially met needs a line saying what is missing.
- **Evidence**: the artefact that proves it: config extract, access review, VAPT report, restore-test result, log-review record. No evidence, no "Met".
- **Last evidence**: the date the evidence was last produced. An old date is a finding.
- **Owner**: the named role answerable for the control on that system.

The three-artefact test per system, run it on a cycle: the last access review, the last VAPT report with closure status, and the log of who accessed a sample record. All three, or the safeguard is asserted, not real.
