Security & Privacy Trust Center
AI Risk Governance is built with privacy-by-design principles at every layer. This page documents the technical security controls, data handling practices, and compliance alignment that protect your organization's data.
Privacy-by-Design Principles
Data Minimization
We collect only the data necessary to deliver governance insights. Search queries are anonymized at the point of collection. Personal identifiers are masked in all administrative interfaces. No raw PII is ever exposed through API responses or downloadable reports.
Purpose Limitation
Data collected for visitor analytics is used exclusively for aggregate platform improvement. It is never sold, shared with third parties, or used for behavioral targeting. Institutional subscribers retain full ownership of their governance data.
Storage Limitation
Automated retention policies enforce strict time-based data lifecycle management. Search telemetry is purged after 30 days. Visitor tracking data is purged after 90 days. No manual intervention is required — daily automated jobs ensure compliance.
Integrity & Confidentiality
Sensitive fields are encrypted at rest using AES-256-GCM. All administrative actions are logged in an immutable audit trail. Server logs are sanitized to prevent accidental PII exposure. Rate limiting protects against abuse and enumeration attacks.
Technical Security Controls
Sensitive data fields are encrypted at rest using AES-256-GCM with PBKDF2-derived keys. Encryption occurs at the application layer before database storage, ensuring data remains protected even in the event of infrastructure compromise.
All personally identifiable information is masked before display in administrative interfaces. Email addresses, names, IP addresses, and company identifiers are redacted using pattern-based masking. Raw PII is never exposed through API responses.
Search queries are hashed using SHA-256 before storage. Only anonymized hashes are retained for aggregate analytics — raw search text is never persisted. Contact information submitted with search requests is redacted at the point of collection.
Visitor tracking data is automatically purged after 90 days. Search telemetry follows a stricter 30-day retention window. Automated daily purge jobs ensure compliance with data minimization principles without manual intervention.
Tiered rate limiting protects all API endpoints against abuse and brute-force attacks. Authentication endpoints enforce strict limits (10 req/min), search endpoints use moderate limits (30 req/min), and general API traffic is capped at 120 req/min per IP.
Every administrative action is recorded in an immutable audit log with user identity, timestamp, action category, and metadata. Audit events cover report access, data exports, retention purges, model management, and system configuration changes.
Strict role separation between standard users and administrators. Administrative procedures are gated by server-side role verification. Tier-based feature access ensures users only see data appropriate to their subscription level.
Server-side log output is sanitized to prevent accidental PII exposure. Email addresses, bearer tokens, API keys, and JWT tokens are automatically redacted before being written to any log stream.
Compliance Alignment
| Framework | Scope | Status |
|---|---|---|
| GDPR | Data minimization, right to erasure, privacy by design | Aligned |
| CCPA | Consumer data rights, disclosure requirements | Aligned |
| SOC 2 Type II | Security, availability, processing integrity | In Progress |
| ISO 27001 | Information security management system | Planned |
| NIST AI RMF | AI risk management framework alignment | Aligned |
Incident Response
In the event of a security incident, our response protocol includes:
Detection & Containment
Automated monitoring, rate limiting, and audit logging enable rapid detection. Affected systems are isolated within minutes.
Notification
Affected institutional subscribers are notified within 72 hours per GDPR requirements, with full incident details and remediation steps.
Remediation & Review
Root cause analysis is conducted for every incident. Findings are documented and controls are updated to prevent recurrence.