Product Roadmap
We build in the open for our enterprise partners. Here is exactly what AI.Guard ships with today, and the capabilities slated for the second half of 2027.
Today
- Detects OWASP Top 10 LLM attacks
- Detects and masks PII and confidential information
- File analysis and handling
- Access control and role-based model routing
- LLM vulnerability scanner
- Connects via proxy, API Gateway, Guardrails SDK, or the unified chat window
H2 2027
- New attack-detection model
- Pre-built detection templates (medical and banking data)
- Data-deletion workflows for regulatory compliance
- RAG oversight
- Kubernetes autoscaling
- Direct on-premises Active Directory / LDAP integration
- Data analyzer for MLSecOps
- Image signing
- Model companion-code oversight
- Model dependency oversight
- Training-artifact oversight