For compliance officers
You are responsible for what the organization is permitted to do with personal data, and for demonstrating it.
Start here
Where most of this role's work happens.
- The governance model — the four enforcement layers
- Manage consent and privacy
- Auditing activity
Your work
| Task | Page |
|---|---|
| Record and fulfil a privacy request | Handle a privacy request |
| Understand what erasure removes and retains | Consent reference |
| Investigate an action | Investigate a change |
| Show how AI access is controlled | AI compliance |
The thing most likely to go wrong
Fulfilling a privacy request against unmerged duplicates.
If the person still exists as several unresolved records, acting on one leaves the others untouched — and you will have told a regulator you erased data you did not.
Search before fulfilling. If duplicates exist, resolve them first. The whole purpose of mastering is that a privacy request has one target.
Erasure retains, and that is correct
Fulfilling an erasure request revokes every consent-based permission and retains those under a basis that does not permit erasure. The response records what was retained and why.
That record is what you show a regulator. Erasing data you are legally required to keep would itself be a violation.
What you can demonstrate
| Claim | Evidence |
|---|---|
| Access is controlled | Four enforcement layers, the innermost in the database |
| Sensitive fields are protected | Masking applied on response generation, so no endpoint leaks them |
| Everything is recorded | Append-only audit log; modification refused, gaps detectable |
| Consent constrains processing | ABAC policies can gate access on an active consent |
| AI is not an exception | Agents act as the user and are subject to identical enforcement |
The strongest of these is the last: it is architectural, not procedural. It does not depend on an agent behaving well.
Next
Last verified against commit 944761c7 (2026-08-03)