Vertex AI Security Guide
Comprehensive security guidance for Google Cloud Vertex AI including model governance, data protection, and enterprise security controls.
Model Security
Secure your ML models throughout their lifecycle
- • Model Registry with versioning
- • Model monitoring and drift detection
- • Explainable AI for transparency
- • Model cards for documentation
Access Control
Fine-grained access management for Vertex AI resources
- • IAM roles for Vertex AI
- • Workload Identity Federation
- • Service account impersonation
- • Resource-level permissions
Data Protection
Encryption
- • Encryption at rest by default
- • Customer-managed encryption keys
- • TLS for data in transit
- • Confidential Computing options
Data Governance
- • Data lineage tracking
- • Feature Store access controls
- • Dataset versioning
- • Data Loss Prevention integration
Monitoring & Auditing
Comprehensive Observability
- • Cloud Audit Logs for all API calls
- • Model Monitoring for performance and drift
- • Prediction request logging
- • Integration with Cloud Monitoring
- • Alerting on anomalies and policy violations