Executive Reporting
Leadership needs a regular view of where the organization is exposed to AI risk, how widely AI is used, and whether policy is being followed. ANTS Platform covers this in three ways: a scheduled executive digest emailed to named people every week or month, a PDF compliance report you generate and share, and dashboards for anyone who wants to look deeper.
Scheduled executive digest
The executive digest is a PDF that summarises policy compliance, AI adoption and exposure across the organization's projects. It is emailed on a schedule, so leadership gets it without signing in.
Set it up in Organization Settings → Executive Digest. The tab appears only for organizations with the Enterprise governance entitlement (see Billing & Plans) and only for users with the Executive digest permission, listed under Governance in the custom role builder. Among the built-in roles, Owner and Security & Compliance can change the digest, and Department Admin and Auditor can view its settings and send history. Your deployment must be able to send email.
Open the tab
Go to Organization Settings → Executive Digest.
Add recipients
Under Recipients, add up to 20 people:
- Add an organization member: choose a member and select Add.
- Add an address outside the org: type the address, select Add and complete step-up verification. Only addresses on your organization's verified domains are accepted.
Choose the schedule
Under Schedule, choose the Frequency (Weekly or Monthly), the Day of the week for a weekly digest, and the Timezone. Leave Include all projects on, or turn it off and select the projects to include.
Turn it on
Turn on Send the executive digest and select Save schedule. The page shows when the next digest will be sent.
Send a test
Under Send history, select Send test to me to receive the latest completed period yourself. Saved recipients do not receive it. You must hold governance read access, the same as a member recipient. You can send one test a minute, after the schedule is saved and with no unsaved changes.
Schedule
| Setting | Behavior |
|---|---|
| Weekly | Sent on the chosen day at 08:00 in the chosen timezone, covering the previous 7 days |
| Monthly | Sent on the 1st at 08:00 in the chosen timezone, covering the previous calendar month |
| Default | Off. The switch stays disabled until at least one recipient is added. |
Each organization has one schedule. Daylight saving time changes do not move the send time. Turning the digest on, or changing its frequency, day or timezone, never sends for a date that has already passed: the first digest goes out at the next scheduled time.
Recipients
| Kind | Who can be added | Step-up |
|---|---|---|
| Member | Organization members who hold governance read access: Owner, Department Admin, Security & Compliance and Auditor among the built-in roles, or a custom role with Governance at Read | No |
| Outside the org | Any address on one of the organization's verified domains. Addresses on other domains are refused. | Yes |
Verify a domain under Organization Settings → AI Access Requests → Company domains. Without a verified domain, only members can be added. A person whose role has no governance access, such as Cost Analyst, can still receive the digest as an address on a verified domain.
Each recipient gets their own email with the PDF attached, so recipients do not see each other's addresses.
Recipients are checked again on every send. Anyone who has left the organization, lost governance read access, has no email address, or whose domain is no longer verified is skipped, and the reason is shown in Send history.
What the PDF contains
- A cover with the organization and the period.
- Policy compliance: agents governed, average coverage, Fully compliant agents (agents that have completed the evidence and attestation mapped to every requirement of their frameworks), open drift alerts, drift alerts raised during the period, and the risk distribution.
- AI adoption: Distinct users, AI tools detected, unsanctioned tools in use, users of unsanctioned tools, and the five most used tools by distinct users. Distinct users counts each user ID once, ignoring case. The Discover AI overview counts each user and device pair, so a person on two devices counts twice there and its figure can be higher.
- Exposure: guardrail detections and requests blocked.
- Framework Compliance Overview for EU AI Act, ISO 42001, NIST AI RMF, SOC 2, ISO 27001 and NIST CSF 2.0, with versions: average coverage, agents Fully Evidenced and requirements per framework.
- Projects: a table of the included projects, then a section for each project with agents, highest risk first, up to 25 projects. When more projects have agents, the cover and the email say how many were left out; they still appear in the table.
- About these figures: how each figure is counted.
The framework figures show how agents' governance evidence maps to each framework's controls. They are not a certification, audit or statement of compliance; the PDF says so under the framework overview and in the footer of every page.
The digest holds aggregated counts only: no prompt or response content, no user names and no hostnames. If a source cannot be read, its block says it "could not be read for this period" rather than showing zero. Adoption and exposure figures follow the dashboards' definitions but are computed separately, so a figure can differ slightly from a dashboard for the same window.
If the PDF is larger than the attachment limit (8 MB by default), it is sent without the per-project sections, and the cover and the email say it is a summary only. If it is still too large, the send fails.
Send history and failures
Send history lists every attempt with when it was sent, its status (Sent, Partly sent, Skipped or Failed), period, number of projects, size, who it was sent to and who was skipped and why. The page also shows a banner when the last attempt failed or was skipped.
| Situation | What happens |
|---|---|
| Email is not configured on the deployment | The send is Skipped. The tab shows a warning. |
| The organization no longer has the Enterprise governance entitlement | The send is Skipped and nothing is sent until the entitlement is back |
| The person who last saved the schedule no longer has permission to change the digest | The send is Skipped until someone with that permission saves the schedule again |
| No recipient is still eligible | The send is Skipped |
| Every email fails, or the PDF cannot be built | The send is Failed and retried at the next hourly check, up to 3 attempts per period |
| Some emails fail | The send is Partly sent and not retried, so nobody gets it twice |
Every change to the schedule, every recipient added or removed and every test send is written to the audit log.
Compliance report
Generate the report on Governance → Audit & Export. Generating it needs the Audit & Evidence Export permission at Read & Write, which Owner and Security & Compliance have among the built-in roles. See Members & Roles.
Open Audit & Export
Go to Governance → Audit & Export, tab Export Reports.
Choose the report
Choose the frameworks, agents and content to include.
Export as PDF
Choose PDF and export. Share the file with the people who need it.
The report covers one project and is built from the governance data at the moment you generate it. To send an organization-wide summary automatically, use the scheduled executive digest.
The PDF compliance report contains:
- Executive Summary: total agents, average coverage and the number of Fully Evidenced Agents.
- Risk Distribution: agents by EU AI Act risk level, as a chart.
- Framework Compliance Overview: average coverage, agents Fully Evidenced and requirements for each selected framework, with a note that the figures are an evidence mapping, not a certification.
- One section per agent with its risk level, coverage and requirement status, evidence details and attestations when selected, and its Tool Governance evidence.
Where each topic is covered
| Topic | In the digest | In the report | Dashboards for more detail |
|---|---|---|---|
| AI exposure | Guardrail detections, requests blocked and the risk distribution | Risk Distribution and each agent's risk level | Risk Dashboard: high, medium and low risk agents and the risk heatmap. AI Violations: PII, PCI, PHI, harmful content and jailbreak detections. AI Tools: tools in use by sanction status. |
| AI adoption | Distinct users, AI tools detected, unsanctioned tools and their users, and the most used tools | Total agents in the Executive Summary | AI Usage Intelligence: active AI users, coverage, sessions, interactions and departments. AI Agent Center: every agent with its requests, cost and status. |
| Policy compliance | Agents governed, average coverage, fully compliant agents, drift alerts and the framework overview with fully evidenced agents | Framework Compliance Overview, fully evidenced agents and requirement status per agent | Framework Coverage: coverage and fully evidenced agents per framework. AI Guardrails: compliant and unprotected agents. Drift Alerts: agents whose behavior no longer matches their evidence. |
| AI spend | Not included | Not included | AI Cost: spend by model, user, agent and environment. |
AI spend is on the dashboards only. Employee adoption figures from Discover AI are in the digest and on the dashboards, not in the compliance report.
Who can see the dashboards
| Dashboard | Built-in roles |
|---|---|
| Risk Dashboard | Owner, Department Admin, Security & Compliance, Auditor |
| AI Usage Intelligence | Owner, Department Admin, Security & Compliance, Cost Analyst |
| Framework Coverage | Owner, Department Admin, Security & Compliance, Auditor |
| Audit & Export reports (download) | Owner, Security & Compliance, Auditor |
For an executive who wants to sign in, Auditor covers the governance dashboards and reports. AI Usage Intelligence needs the Adoption Overview permission, which a custom role can add.
Next steps
- Audit & Export: on-demand PDF and Excel reports.
- Risk Dashboard: how risk is computed.
- AI Usage Intelligence: how adoption is counted.