# Security

AI security controls engineered into the request path, plus framework alignment.

## Control path
Decode → Classify → Redact → Route → Log.

## Controls
- Prompt-injection defense
- PII & secret redaction
- Insecure-output handling
- Immutable, hash-linked audit trail
- Zero-trust request routing
- Least-privilege provider credentials

## Framework alignment
- OWASP LLM Top 10 — aligned
- NIST AI RMF — aligned
- KVKK / GDPR — aligned
- SOC 2 Type II — audit in progress (not yet certified)
- ISO/IEC 27001 — audit in progress (not yet certified)

Alignment reflects controls engineered to each framework. SOC 2 Type II and ISO/IEC 27001 audits are in progress — this is not a claim of certification.