Content Governance & Moderation FAQ
Answers to common setup and operating questions about Content Governance and Content Moderation.
For what these surfaces are and how to set them up from scratch, see the
Content Governance & Moderation Overview.
Setup
Do I need to enable governance and moderation separately?
Yes. Content Moderation is controlled by the moderation type setting at Admin → General Settings → Content Moderation → Settings. The default is Moderate only flagged content (users report content manually; nothing is screened automatically). Content Governance is enabled per app — toggle Enable Auto Governance on each app’s card at Admin → General Settings → Content Governance → Settings. Neither surface requires a separate app licence; both are available to every admin.
Who can review the moderation queue?
Admins and super admins always can. If the Team admins can moderate setting is on (it is by default), team admins can also review items that belong to their team. An assigned reviewer on the source content can review that specific item.
Which content surfaces are screened by moderation?
Eleven policies are registered: Ask AI, Plugin Builder, Forms Builder, Forms Submissions, Candidate Portal, Workforce Bots Portal, Broadcast Posts, Broadcast Comments, Recognition Posts, Chat Messages, and Plugin Records (App Builder custom apps). Each policy decides which checks apply to its surface — for example, Ask AI checks for prompt injection while Broadcast Posts do not.
Which apps have governance adapters?
Fourteen app slugs are registered across 16 adapters: Libraries, Sites, Policy Hub, SOP Hub, Safety Hub, Wikis, Mango IQ, Mango GTM, Workforce Bots, Workspace, Drive, Inspections, Communications (two adapters — News Feed and Comms Hub), Contracts, and Onboarding Hub. Only apps with a registered adapter appear on the governance settings page.
Permissions and access
Why can’t a team admin see the moderation queue?
The queue requires the Team admins can moderate setting to be on. Check Admin → General Settings → Content Moderation → Settings. If it is on, the team admin can only review items from their own team — items from other teams or with no team context will not show a review option.
Are admins themselves screened by moderation?
No. Admins and super admins are always exempt from screening, regardless of the moderation type or user scope setting.
Day-to-day
What happens when I approve or reject a moderation item?
Approve publishes the held content and records the reviewer and timestamp. Reject removes the content. In both cases, if the Notify authors on decision setting is on (default: on), the author receives an email telling them the outcome.
What is the “unscreened publishes” panel below the moderation queue?
When an automated screen does not return in time, the system fails open — the content publishes without being checked. These events are recorded as their own kind of finding: the content was allowed through, and no review decision is attached to it. They appear in a separate panel titled Published without an automated check (up to 10 rows, covering the last 7 days) because they are not decisions awaiting review — they are control-gap records for operational awareness.
How often does the governance scan run?
The weekly scan runs every Monday at 7:00 AM UTC. You can also trigger a manual scan from the Findings page; there is a 5-minute cooldown between scans.
How often does gap analysis run?
The gap analysis job runs every Wednesday at 7:00 AM UTC, separate from the governance scan. It requires gap analysis to be enabled for at least one app, and needs at least 50 search log entries in the last 30 days before it will produce results.
What does snoozing a governance finding do?
Snoozing defers a finding until a date you choose. The finding stays Open (so the scanner does not re-detect and duplicate it) but drops out of the active worklist. It reappears automatically when the snooze date passes. The snooze date must be in the future.
Can governance findings auto-resolve?
Yes. Each weekly scan re-evaluates open findings. If the underlying condition has cleared — a stale item was updated, a missing source was restored, a custom adapter check no longer fires — the finding is resolved automatically. If the subject record has been deleted entirely, the finding also auto-resolves.
When something looks wrong
“I enabled AI moderation but nothing is getting blocked”
Check the AI Mode setting. When AI moderation is first enabled, it starts in Observation mode, which logs what the AI would flag without blocking or holding anything. Switch to Live mode at Admin → General Settings → Content Moderation → Settings when you are satisfied with the detection quality.
“The governance scan shows zero findings for most apps”
Each app’s governance adapter must be individually enabled. Go to Admin → General Settings → Content Governance → Settings and check that Enable Auto Governance is toggled on for each app you want scanned. An app that is not toggled on is skipped entirely during the scan.
“A governance finding has been open for weeks — why hasn’t it escalated?”
Auto-escalation only applies to Low and Medium severity findings. Low findings escalate to Medium after the configured number of days (default 30); Medium findings escalate to Critical after another window (default 60 days). Findings that an adapter originally rated High are deliberately excluded from the escalation ladder — they stay at High unless you manually change them. Also, only Open and Acknowledged findings are escalated; Resolved and Dismissed findings are not.
Licensing and limits
Are there limits I should know about?
| What | Limit | Where it applies |
|---|---|---|
| Review interval (governance) | 7–365 days per app | Review Interval (days) on each app’s card in Content Governance settings |
| Escalation windows | 1–365 days each | Severity Auto-Escalation on each app’s card in Content Governance settings |
| Default review interval (Libraries) | 90 days | The Libraries governance adapter, before you change it |
| Default low → medium escalation | 30 days | The weekly governance scan, before you change it |
| Default medium → critical escalation | 60 days | The weekly governance scan, before you change it |
| Manual scan cooldown | 5 minutes | Run Scan on the Content Governance findings page |
| Broken links checked per scan | 50 max per run | The broken-link check inside each governance scan |
| Broken link timeout | 10 seconds | The broken-link check inside each governance scan |
| Broken link failure threshold for critical | 3 consecutive failures | The broken-link check inside each governance scan |
| Content excerpt stored per moderation finding | 500 characters | Each item in the moderation queue |
| AI negativity tolerance range | 0–100 (0–25 strict, 26–50 moderate, 51–75 relaxed, 76–100 lenient) | Negativity Tolerance in Content Moderation settings |
| Gap analysis minimum search logs | 50 in 30 days | The weekly gap analysis run |
| Unscreened-publish panel window | 7 days | The Published without an automated check panel on the moderation queue |
More help
- Content Governance & Moderation Overview
- Audit Logs & Activity History — tracks moderation decisions and governance changes
- Ask AI — the assistant answers Content Governance & Moderation questions from these articles.