Security Overview

Controls include tenant isolation, RBAC, hashed API and SCIM tokens, redacted logs, webhook signing, KMS-backed secret storage options, SAML SSO, SCIM v2 provisioning, audit exports, and SIEM delivery.

For AI-specific risk reduction, the gateway can detect, block, redact, monitor, and report prompt injection, secrets, PII, unsafe outputs, system prompt leakage, RAG poisoning, and grounding failures.

Security is operated as defense-in-depth. No public page should be read as a promise of complete protection.