Skip to main content

Agent Compliance

The Compliance tab on an agent's detail page shows that agent's regulatory posture and the evidence behind it. For organization-wide compliance, see Compliance & Audit.

The tab has four sub-tabs: Overview, Evidence exports, Findings, and History.

Framework coverage

Each agent's controls are mapped to three frameworks:

FrameworkRequirement unit
EU AI Actobligations
ISO/IEC 42001controls
NIST AI RMFoutcomes

For each framework the Overview reports a posture score, Covered / Partial / Failing counts, and the number of mapped requirements. Aggregated across frameworks it shows an overall agent posture % and a posture trend over 7d / 30d / 90d / 1y.

Findings

Open findings are grouped by severity — Critical / High / Medium / Low. A Top remediation list ranks actions by their score impact (e.g. "remediate failing resources for EU AI Act Art. 10"). A requirement status grid shows Covered / Partial / Failing / N/A per framework.

Actions

ActionDescription
Export evidence packGenerate a downloadable evidence bundle for the agent's mapped requirements.
Re-evaluateRe-run the compliance evaluator. The evaluator version is shown (e.g. 2026.05.01-v1).

Recent activity logs attestations, exports, and signal changes.