Prevent agents or models from executing actions outside their permitted scope. Define enforcement rules with conditions, thresholds, and automated actions. Every enforcement decision is recorded in the Provenance Vault.
Define, enforce, and track your governance rules.
Define custom governance rules with conditions, thresholds, and automated actions.
Rules evaluated on every AI output in real time with automatic enforcement.
Complete audit trail of every rule violation with severity classification.
Create governance rules through the visual builder or the API. Either way, enforcement is immediate.
Set conditions, thresholds, and actions via API or visual builder.
Rules go live immediately across all AI outputs in your workspace.
Real-time violation tracking with email alerts and audit trail.
{
"rule_id": "rl_pii_block",
"name": "Block PII in Outputs",
"condition": {
"type": "contains_pii",
"categories": ["ssn", "credit_card", "phone"]
},
"action": "block",
"severity": "critical",
"alert": {
"email": ["compliance@company.com"],
"webhook": "https://..."
}
}From manual review to automated enforcement.
Inconsistent
Xybern Custom Rules
Full control over every aspect of your governance rules.
Set trust score minimums, bias limits, and PII sensitivity levels.
Block, flag, alert, or redirect outputs based on rule conditions.
Email notifications, webhook callbacks, and Slack integrations.
Time-based rules for business hours, blackout periods, and review windows.
Every rule violation tracked with full context and severity classification.
Rule: All Rules
PII Detected in Output
Criticalrl_pii_block
Trust Score Below Threshold
Highrl_min_trust
Bias Pattern Detected
Mediumrl_bias_check
Unauthorized Data Access
Criticalrl_data_gate
Use POST /api/v1/rules to define custom governance rules programmatically. Set conditions, actions, and alert recipients in a single request.
Rules are activated immediately upon creation. Every rule violation is tracked, logged, and available through the violations API endpoint for enforcement reporting and provenance.
// POST /api/v1/rules
{
"name": "Minimum Trust Score",
"condition": {"trust_score_below": 0.7},
"action": "block",
"alert": ["admin@company.com"]
}
// Response
{
"rule_id": "rl_min_trust_7x",
"status": "active",
"created_at": "2025-01-15T10:30:00Z"
}Deploy custom governance rules across every AI system in your organisation.