Privacy / Data Protection Assessment: FortProof
|
|
| Status |
Input for your data protection officer. Does not claim compliance with UK GDPR, the Data Protection Act 2018 or any other law. Legal and DPO review is required. |
| Owner |
Programme owner, with the DPO |
| Companion |
DPIA considerations, Data Retention Policy |
Yes, in limited categories. It is primarily a security-assurance system about applications, but it necessarily processes data about the people who build, own and assess them.
| Category |
Examples |
Data subjects |
Source |
Necessity |
| Identity of owners and roles |
Names, work emails, identity-provider IDs, GitHub or GitLab logins of business/technical owners, champions |
Staff, contractors |
application.yaml, the identity provider, GitHub or GitLab |
Required for ownership and escalation |
| Workflow actors |
Requester/approver of exceptions, reviewers, assessors |
Staff |
Platform sessions |
Required for SoD and accountability |
| Audit records |
Actor, action, time, source IP, user agent |
Platform users |
API |
Required for security of the platform |
| Developer activity inside raw artefacts |
Commit author names/emails in scanner output, committer of an exposed secret |
Developers, external contributors (public repos) |
Scanner output |
Incidental. Minimised in normalised records. |
| Training completion |
Course completion dates |
Engineers |
LMS export |
Team-level coverage only |
| Security findings |
Vulnerabilities in applications |
Not personal data in themselves |
Scanners |
Core purpose |
| Application data |
Customer or end-user data processed by the applications |
— |
Not processed. The platform never ingests application data. |
— |
2. Purposes
Security assurance of applications · risk management and remediation tracking · audit of security controls · platform security (audit logging). Explicitly excluded purpose: monitoring or evaluating individual employee performance. Per-person metrics are not built, and team-level reporting is the minimum granularity.
3. Data protection by design measures
| Measure |
Implementation |
| Minimisation |
Secret values are never stored. Committer emails are stripped from normalised findings. Only the attributes needed for ownership come from the identity provider. |
| Purpose limitation |
Prohibited-use statement in the charter and acceptable-use notice. No individual leaderboards. |
| Access control |
RBAC + RLS. Auditor access time-bound. Admins lack default detail access. (Access Control Model) |
| Retention |
Defined per category. Leavers pseudonymised after 90 days. (Retention Policy) |
| Security |
Private networking, encryption in transit/at rest, immutable audit, threat model (platform security) |
| Transparency |
Privacy notice for platform users and engineers (to draft with the DPO) |
| Data location |
Azure UK South (primary), UK West (backup/DR). No transfer outside the UK expected. Confirm for Sentinel, GitHub and Jira, which are SaaS. |
4. Points requiring DPO / Legal determination
- Lawful basis for processing staff data in this platform (likely legitimate interests or public task, depending on organisation type). This needs a DPO determination; this document does not decide it.
- Whether processing of external contributor data in raw artefacts of public repos needs specific notice.
- Whether WORM retention (data that cannot be erased early) is acceptable for incidental personal data, with the minimisation measures above.
- Whether a full DPIA is required. See the screening in DPIA considerations. A conservative reading suggests one is advisable, given systematic processing linked to employees' work.
- Employee consultation or works-council requirements, if applicable.
- Data sharing with the SOC (Sentinel) and Jira: confirm these are within existing processing records.