redwitch

Threat Model for Menstrual-Cycle Tracking Data

Framework: STRIDE (security threats) + LINDDUN (privacy threats) Scope: Menstrual cycle logs, symptoms, sexual activity, fertility windows, partner-sharing, device metadata, identifiers, geolocation, backups, analytics events, notification content, and user account data.


1. STRIDE Threat Model (Security-Focused)

S — Spoofing Identity

Threats

Root Causes

Potential Harms

Mitigations


T — Tampering

Threats

Root Causes

Potential Harms

Mitigations


R — Repudiation

Threats

Root Causes

Potential Harms

Mitigations


I — Information Disclosure

Threats

Root Causes

Potential Harms

Mitigations


D — Denial of Service (DoS)

Threats

Root Causes

Potential Harms

Mitigations


E — Elevation of Privilege

Threats

Root Causes

Potential Harms

Mitigations


2. LINDDUN Threat Model (Privacy-Focused)

L — Linkability

Threats

Harms

Mitigations


I — Identifiability

Threats

Harms

Mitigations


N — Non-Repudiation (In Privacy)

(In privacy models, this refers to inability to plausibly deny an association with sensitive data.)

Threats

Harms

Mitigations


D — Detectability

Threats

Harms

Mitigations


D — Disclosure of Information

(Overlaps with STRIDE “Information Disclosure,” but from a privacy lens.)

Threats

Harms

Mitigations


U — Unawareness

Threats

Harms

Mitigations


N — Non-Compliance

Threats

Harms

Mitigations


3. High-Risk Scenarios (Cross-STRIDE/LINDDUN)

These represent the most severe, multi-category threats:

A. Intimate Partner Violence (IPV) Coercion

B. Reproductive-Rights Criminalization

C. Third-Party SDK Leakage

D. Data Breach of Cloud Storage

E. App Shutdown or Acquisition


4. Summary of Mitigation Priorities

Based on severity and likelihood:

  1. Local-first architecture with strong encryption.
  2. Eliminate or drastically limit third-party SDKs.
  3. Stealth and safety modes for IPV contexts.
  4. Granular, revocable consent with no bundled permissions.
  5. Privacy-preserving analytics or zero analytics.
  6. Offline access and minimal server logging.
  7. Inclusive design for gender-diverse and vulnerable users.
  8. Data retention limits and “zero-knowledge” policies where possible.