Built to pass your security review
PhishGuard handles the most sensitive signal an organization has — how its people respond to attack. This page describes, plainly, how the platform protects that data. Every statement here reflects how the system actually works; if your security team wants to go deeper on any of it, talk to us and we'll walk through the specifics.
Tenant isolation
PhishGuard is multi-tenant with hard boundaries. Every record is scoped to its organization, and a dedicated guard layer audits and enforces organization scoping on database queries at runtime. Public, unauthenticated endpoints serve platform-level aggregate figures only — never anything derived from tenant data. A federated sign-in asserting an identity from another tenant is denied outright.
Authentication
Password sign-ins are protected by multi-factor authentication with one-time codes and account lockout on repeated failure. Organizations can bring their own identity provider via per-tenant SAML single sign-on: assertions must be signed by the tenant’s registered certificate, replay is rejected, and the browser handoff uses short-lived single-use codes so session tokens never appear in URLs. MFA challenge data likewise never travels in URLs.
Credentials & secrets
User passwords are stored only as adaptive one-way hashes. Integration API keys are org-scoped, shown once at creation, stored as SHA-256 hashes, and individually revocable. The breach-exposure monitoring module is built around a strict redaction boundary: recovered credential values from breach or infostealer corpora are read only to set exposure flags and are never written to our systems — only field names survive ingestion.
Transport security
All traffic is encrypted in transit. The platform accepts TLS 1.2 and 1.3 only, with a restricted modern cipher suite, and sends HTTP Strict Transport Security with a long max-age. Standard hardening headers (content-type sniffing protection, referrer policy, frame protections) are served platform-wide.
Auditability
Security-relevant events are recorded in an audit log: sign-ins and failures (including the authentication method, so SSO logins are attributable), administrative changes, MFA challenges, and authorization denials. Email deliveries are logged with outcome. Audit data is available to your administrators for review and export.
Abuse prevention
Authentication and public endpoints are rate-limited per source. Account-recovery endpoints return the same response whether or not an account exists, preventing account enumeration. The public demo-request form combines rate limiting with honeypot detection.
Operational practice
Deployments are containerized with the database isolated from public exposure. Changes ship with tagged rollback images so any release can be reverted in minutes. Security review is part of feature work — capabilities that touch authentication or tenant boundaries are tested adversarially before they are enabled for customers.
Reporting a vulnerability
If you believe you've found a security issue in PhishGuard, we want to hear about it. Contact us through the contact form or through your account manager, and we'll acknowledge your report promptly and keep you informed as we investigate. We ask that you avoid accessing other tenants' data while testing.