Implementation Layer
AI Governance Controls
Operational controls for real-world enterprise AI systems — organized by domain, mapped to regulations, with maturity levels and implementation guidance.
Human Oversight
Review gates, approval workflows, and override mechanisms for AI decisions.
6 controls
AGTAgentic AI
Goal constraints, action boundaries, and escalation paths for autonomous AI agents.
8 controls
SECSecurity
Adversarial input defense, prompt injection protection, and model access controls.
5 controls
ALCAudit & Logging
Immutable records of AI decisions, inputs, outputs, and model versions.
5 controls
CHMChange Management
Model release governance, version rollback, and change approval workflows.
5 controls
DGCData Governance
Training data provenance, privacy controls, and data retention policies.
5 controls
MONMonitoring & Drift
Performance drift detection, anomaly alerting, and operational dashboards.
5 controls
SAFSafety & Reliability
Graceful degradation, fail-safe defaults, and reliability under adversarial inputs.
5 controls
IRCIncident Response
Containment, investigation, and remediation procedures for AI system failures.
5 controls
PRCProcurement
Third-party AI vendor due diligence, contractual obligations, and offboarding.
5 controls
6 controls matching filters
Agentic AI
3 controlsAgent Permission Boundaries
Apply least-privilege principles to AI agents by explicitly defining and enforcing the tools, APIs, data sources, and actions each agent is authorized to access.
Multi-Agent Trust Hierarchy
Define explicit rules for which agents can instruct, invoke, or delegate authority to other agents in multi-agent systems.
Agent Environment Isolation
Run AI agents in isolated execution environments that limit their ability to access host systems, network resources, or data beyond what their task requires.