Security & Identity
AI Security Gateway
Security-scanned AI gateway chat: on the gateway chat surface, prompts pass the AI security input scanner and redaction before reaching a model, model output passes the output sanitizer before it is returned, org AI governance policy is evaluated per request, and blocked requests are audit-logged.
Outcomes
- • Input scanning with redaction on the gateway chat surface (available)
- • Output sanitization before responses are returned (available)
- • Org AI governance policy evaluation per request (available)
- • Security audit events for blocked requests (available)
Capabilities
- • Input scanning with redaction on the gateway chat surface (available)
- • Output sanitization before responses are returned (available)
- • Org AI governance policy evaluation per request (available)
- • Security audit events for blocked requests (available)
- • Org model allowlist enforcement on the gateway (available)
- • OWASP LLM Top 10 risk controls (planned)