Transparency

Methodology & Governance

Rigorous, reproducible standards for data acquisition, harm quantification, and ethical assessment. All frameworks are version-controlled and publicly auditable.

1. Data Acquisition Protocols

All research and framework inputs are derived through three validated channels. We never use scraped PII or non-consensual data sources.

  • Official APIs: Primary source for platform transparency reports, government open data portals, and regulatory filings. All API access complies with provider terms of service.
  • Structured Web Collection: Used only for publicly available, non-personal aggregate data (e.g., proxy pricing pages, public court records). Automated collection respects robots.txt and implements rate limiting to prevent service disruption.
  • Direct Mission Submissions: Anonymized case data provided by diplomatic partners under bilateral data sharing agreements. All submissions are encrypted in transit and at rest.

Validation: Every dataset undergoes cross-source verification before inclusion in any framework or report. Single-source claims are explicitly flagged as preliminary.

2. Harm Quantification Matrix

The Global Harm Framework uses a six-tier severity scale applied across eight impact domains. Scores are derived from documented evidence, not subjective assessment.

  • Evidence Threshold: Minimum of two independent corroborating sources required for any harm classification above Level 2.
  • Cross-Border Weighting: Harms affecting individuals outside their country of citizenship receive a 1.5x multiplier when jurisdictional gaps prevent domestic remedy.
  • Status Neutrality: As of v2.1, all visa and residency statuses are weighted equally. Previous versions differentiated between citizen and non-citizen impacts; this was removed following Round One diplomatic consultations.

3. Ethical Sourcing Assessment Criteria

Used to audit proxy network supply chains. Compliance requires meeting all five mandatory criteria:

  • Informed Consent: End users must actively opt-in with clear, non-deceptive language. Pre-checked boxes and bundled consent are automatic failures.
  • Compensation Transparency: Users must receive fair value exchange (monetary or service-based) proportional to bandwidth contributed.
  • Traffic Restrictions: Networks must block known fraud, abuse, and illegal content categories at the infrastructure level.
  • Audit Trail: Providers must maintain immutable logs of consent timestamps and user withdrawal events for minimum 24 months.
  • Third-Party Verification: Annual independent audit report published within 90 days of fiscal year-end.

4. Data Governance & Privacy Policy

Transparency-X operates under a strict minimization principle. We do not collect, store, or process personally identifiable information except where explicitly required for authenticated diplomatic services.

  • Diplomatic Portal Access: Mission codes are hashed using bcrypt. Session tokens are HTTP-only, secure-flagged, and expire after 8 hours of inactivity.
  • Contact Form Submissions: Retained for 90 days for response tracking, then permanently deleted. Never used for marketing or shared with third parties.
  • Analytics: Cloudflare Web Analytics collects only anonymized, aggregated metrics. No cookies, no fingerprinting, no cross-site tracking.
  • Data Residency: All infrastructure hosted on Cloudflare’s EU region unless otherwise specified in bilateral agreements.

5. Global Harm Framework Changelog

Version Date Key Changes
v2.1 Jun 9, 2026 Removed citizen/non-citizen weighting differential; added cross-border jurisdictional gap multiplier; integrated proxy ethics criteria into harm scoring; expanded Brazilian mission consultation feedback loop.
v2.0 Mar 15, 2026 Introduced six-tier severity scale; added eight impact domains; replaced binary harm/no-harm classification.
v1.2 Nov 2, 2025 Added evidence threshold requirements; introduced single-source flagging protocol.
v1.0 Jul 20, 2025 Initial release. Binary harm classification; single-domain assessment.

All prior versions remain accessible via our research archive for reproducibility purposes.