Framework Coverage
Framework Coverage is the compliance scoreboard. It rolls up approved evidence into a coverage percentage per framework and shows exactly which controls are complete, partial, or not yet assessed — so you know how close you are to audit-ready and where the gaps are.
Enterprise feature. Open it from AI Governance → Compliance → Framework Coverage. Coverage reflects approved evidence — see Requests & Review.
What it shows
For each framework — EU AI Act, NIST AI RMF, ISO 42001 — you see:
- Coverage % — completed controls as a share of total required controls.
- Control breakdown — how many controls are completed, partial, or not assessed.
- The gap list — the specific controls still outstanding.
- Agent × framework status — which agents meet which framework requirements, so you can spot the laggards.
Coverage status
Coverage rolls up into an at-a-glance status per framework:
| Status | Meaning |
|---|---|
| Pass | High coverage — effectively audit-ready |
| Monitoring | Partial coverage — in progress, watch the gaps |
| Fail | Low coverage — significant work remains |
How coverage is calculated
Coverage is the ratio of completed evidence items to total required items for the framework, across the agents in scope. Approving evidence in Requests & Review is what moves the needle; runtime-derived fields contribute automatically as your agents run.
Filter the view to a single framework to focus a remediation push (e.g. "get EU AI Act to Pass"), or view all frameworks together for an executive summary.
Closing gaps
- Sort by coverage to find the framework or agent that's furthest behind.
- Open the gap list to see the outstanding controls.
- Create or assign the evidence for those controls in Evidence Collection.
- Approve the submissions — coverage updates as evidence is approved.
Next steps
- Evidence Collection — fill the outstanding controls.
- Drift Alerts — make sure approved controls stay true at runtime.
- Audit & Export — package coverage for auditors.