--- sourceDocument: Brazil Enable AI sourceDocumentLink: https://www.servicenow.com/docs/r/intelligent-experiences Release : - brazil ft:locale : - en-US ft:publication_title : - Brazil Enable AI ft:clusterId : - platai bundleId : - platai workflow : - Platform --- # Governance insights # Detecting governance insights {#ariaid-title1} Release version: Brazil Updated October 1, 2026 ![](https://www.servicenow.com/docs/portal-asset/ico-clock) 3 minutes to read Scheduled detection jobs evaluate your AI portfolio and generate the Stalled Lifecycle Tasks and Significant Issues insights that appear as recommendations. The Your top recommendations widget surfaces two insights that identify governance work needing attention across your AI portfolio: Stalled Lifecycle Tasks and Significant Issues. Both run as scheduled detection jobs rather than as on-demand reports, so the widget reflects the most recent scheduled run rather than real-time state. {#gov-airc-insight-detection__table_insight_purpose__entry__2} | Insight | Purpose | |-|-| | Stalled Lifecycle Tasks | Identifies lifecycle tasks on AI systems that have seen no progress or updates and require attention to resume completion. | | Significant Issues | Identifies open issues on AI systems that have a meaningful impact on risk or compliance and warrant timely review. | [Table 1. Governance insights] {#gov-airc-insight-detection__table_insight_purpose} These insights provide the following benefits: * Surface high-priority governance work without requiring you to search across individual AI systems. * Detect stalled work and significant issues on a recurring schedule. * Provide a portfolio-wide, severity-prioritized view of items that need attention. {#gov-airc-insight-detection__ul_insight_benefits} ## How each insight is determined {#gov-airc-insight-detection__section_how_determined} Both insights follow the same two-stage pattern. First, a detection job identifies AI systems and records that match a set of criteria. Then a generative AI skill prioritizes which matches are most important to surface. If the AI skill call fails or returns no results, the detection job falls back to a deterministic, rule-only ranking so the insight still generates. For the exact criteria, detection intervals, and fallback ranking for each insight, see [Governance insight detection reference](https://www.servicenow.com/docs/cqGvWDYixTkLyTEPcMmrkA "Thresholds, criteria, and refresh schedule that determine when an AI system appears in the Stalled Lifecycle Tasks or Significant Issues insight."). Warning: AI-generated summaries and prioritization can be inaccurate or incomplete. Review each recommendation and the records behind it before you act on it. ## Automating a response {#gov-airc-insight-detection__section_automate_with_otto} Stalled Lifecycle Tasks recommendations include an Automate with Otto action. Selecting it opens a confirmation dialog box with the message Automate this action item? This will initiate a workflow run by an unsupervised agent. Review the recommendation before you proceed. Select Automate to proceed, or select Cancel. Significant Issues recommendations don't have this option. Resolve them directly from the issue list. When you confirm, the agent sends an email reminder to the assignee of each overdue or dormant lifecycle task on the AI systems in the recommendation. Each person receives one email per AI system that lists their tasks. The agent doesn't complete, reassign, or update the tasks. The recommendation shows the outcome of the run, including who was notified about which tasks. ## Viewing recommendation details {#gov-airc-insight-detection__section_view_details} Selecting View details on a recommendation in the Your top recommendations panel opens a record with the fields in the following table. {#gov-airc-insight-detection__table_recommendation_record__entry__2} | Field | Description | |-|-| | Short description | Generated summary of the recommendation. The summary can also name specific AI systems and recommend actions, not just report totals. * For a Significant Issues recommendation, this includes the number of active issues found, managed AI assets affected, and top issues and assets shown. For example: 11674 active issues across 483 managed AI assets require review. Showing the top 50 issues on the 10 highest-priority assets. * For a Stalled Lifecycle Tasks recommendation, this includes the total number of overdue tasks, dormant tasks, and governed AI assets affected. For example: 319 overdue and 473 dormant lifecycle tasks across 213 governed AI assets. {#gov-airc-insight-detection__ul_prc_xwm_tkc} | | Activity type | AI recommendation for these insights. | | Type | Insight category: either Stalled Lifecycle Tasks or Significant Issues. | | Initiated date | Date and time the detection job generated this recommendation. | [Table 2. Recommendation record] {#gov-airc-insight-detection__table_recommendation_record} The record also includes the same action and a History section. For a Significant Issues recommendation, the default action is Open, with Complete and Delete available from the drop-down list. For a Stalled Lifecycle Tasks recommendation, the default action is Automate with Otto, with Complete and Delete available from the drop-down list. There is no separate Open action for this insight. For a Significant Issues recommendation, selecting Open navigates to a filtered Issue list scoped to that recommendation's issues. Results are sorted by priority, then by how long each issue has been open. ## Viewing insights {#gov-airc-insight-detection__section_where_shown} Both insights appear in the Your top recommendations widget, which is shown both on the AI Control Tower home page and on the Risk \& Compliance tab under Govern. Select See all Recommendations in Activity Center to see the full list of assigned and unassigned action items generated by both insights. For more information, see [Act on AI asset governance insight recommendations](https://www.servicenow.com/docs/95VrE9rFLA_qLKBTj52Vag "Resolve Stalled Lifecycle Tasks and Significant Issues recommendations to keep AI systems moving through their governance life cycle.").