Insights and Governance Reports

Use aggregate adoption and risk views to identify where investigation should begin, then drill into underlying users, AI assets, accounts, and findings.

Navigation: Insights > AI Insights or Governance Reports.

Configuration and options: AI Insights exposes a date range plus adoption, department, sensitive-data, user, app, category, and approval-status widgets. Governance Reports exposes Summary Dashboard, Incident Table, and Executive Summary.

Role-specific value: Administrators use trends and governance summaries for program oversight; Engineers use drill-down links and incident filters to build an investigation queue.

Verification: A dashboard observation can be reproduced in the underlying User, AI Asset, Account, or Finding records for the same time range.

Populated AI Insights overview with adoption, department, sensitive-data, and user widgets.

AI Insights walkthrough

1. Set the date range before comparing widgets.

2. Review AI Adoption Trend for changes in participating users and AI services or assets.

3. Use Usage by Department and View All Users to identify who is driving activity.

4. Use Sensitive data classification and View All Findings to pivot from category counts to evidence.

5. Review Top Users, Top AI Apps, Categories of AI Apps, and approval status; pivot to AI Assets for current inventory detail because Applications is end-of-life.

6. Treat “Not enough data” as a data-availability condition for the selected window, not a zero-risk result.

Populated Governance Reports Summary Dashboard with justification and bypass trends.

Governance Reports walkthrough

1. Summary Dashboard: choose date range and granularity; review justification trend, total bypass events, key metrics, sensitive-data breakdown, domain activity, and user response.

2. Incident Table: search by email/finding ID and filter risk level or justification status for the selected date range.

3. Executive Summary: use the generated narrative as a reporting aid, then validate its statements against dashboard metrics and underlying incidents.

Exports: The guide does not establish a direct Export control on AI Insights or Governance Reports. Use supported Users, Findings, or Audit Log exports to preserve reproducible evidence.

Expected result: The team can explain which widget triggered the investigation and reproduce it in detailed records.

Verification: Keep the same date range while pivoting, then confirm users, AI assets, classifications, and findings reconcile with the aggregate view.

Use the end-to-end workflow to investigate a signal from these reports.