Personal Exposure Defense
A repeatable, governed loop that turns raw exposure into a defensible, audit-ready posture.
What it is
Personal Exposure Defense (PED) is the discipline of finding where a protected principal is exposed in public and public-record surfaces, verifying that exposure with evidence, scoring its real-world risk, and reducing it through governed remediation. Cloak Harbor operationalizes PED as a control plane rather than a one-off scan.
The operating loop
Every engagement follows the same ten stages. The demo walks one synthetic executive principal through all of them.
1 — Intake
Create a consent-scoped principal profile: aliases, family links, locations, business entities, identifiers, devices, communication preferences, risk tolerance and approval rules. Intake records whether the case uses synthetic or authorized data.
2 — Exposure discovery
Curated findings are surfaced across people-search-style records, property clues, public-record-style records, corporate filings, social references, credential exposure, utility clues and identity linkage. The demo never runs an unbounded live crawl of real people.
3 — Evidence validation
- Source URL and source class
- Capture timestamp and screenshot / evidence target
- Confidence and exploit path
- False-positive status and why it matters
4 — Risk scoring & StealthScore
The SigReduce rubric scores severity, confidence, exploitability, family exposure, physical proximity, fraud utility, social-engineering leverage and recurrence likelihood. These roll up into a StealthScore and a projected remediation delta, labeled as demo calibration.
5–6 — Operations & cadence
Findings move through a Kanban board — New exposure, Validated, Action ready, Waiting partner, Closed / monitor — while a calendar schedules opt-out deadlines, partner review, client approval, re-scan and report-out.
7–8 — Agent-assisted action & human approval
AI drafts takedown notices, privacy requests, mitigation recommendations and analyst notes in dry-run mode. Nothing is submitted externally until a human approves. AI output is always labeled and never treated as final truth.
9–10 — Trace, report & recurrence
A one-page executive report captures before/after status, open risks, completed and pending actions, recurrence watch and the next 30 days — backed by a trace evidence timeline and ongoing monitoring.