AI governance by function
Function-led planning helps teams identify where AI Workers can draft, classify, summarize, route, prepare evidence, or support decisions while preserving human accountability, review gates, and lifecycle controls.
Use this page when the team already knows the business function or workflow owner. The function lens helps clarify business outcome, accountable owner, workflow boundary, data and tool assumptions, human review path, evidence expectations, and the next governed step into Agent Builder, Passport planning, or advisor support.
Function lenses
Each function should start with work visibility, human ownership, data boundaries, decision impact, evidence needs, and the handoff path into the Scaled Agents operating model.
Plan AI Worker support for case triage, knowledge retrieval, response drafting, service recovery, escalation routing, quality review, and customer-experience reporting.
Clarify customer-impact boundaries, approved knowledge sources, prohibited actions, review gates, and service evidence before external communication.
Plan AI Worker support for analysis preparation, reporting support, reconciliation assistance, controls evidence, exception tracking, and planning narratives.
Clarify data sensitivity, approval paths, segregation of duties, audit evidence, and what remains owned by finance reviewers.
Plan AI Worker support for control mapping, evidence preparation, issue intake, policy review support, framework crosswalks, and remediation tracking.
Clarify reviewer ownership, source-of-truth boundaries, exception handling, audit trail needs, and non-certification language.
Plan AI Worker support for alert summarization, incident preparation, evidence capture, connector review, access-boundary checks, and security posture reporting.
Clarify least privilege, sensitive-data controls, prompt-injection risk, tool permissions, human escalation, and incident response ownership.
Plan AI Worker support for discovery synthesis, backlog refinement, acceptance criteria, release readiness, defect triage, test planning, and delivery evidence.
Clarify model/provider assumptions, runtime boundaries, change control, rollback expectations, observability, and ownership across lifecycle stages.
Plan AI Worker support for status synthesis, RAID tracking, decision records, dependency visibility, meeting preparation, and governance-ready delivery evidence.
Clarify decision ownership, escalation paths, source trust, portfolio visibility, value tracking, and review cadence.
These pages are planning aids. They do not approve production use, legal conclusions, compliance conclusions, security authorization, live integrations, or autonomous approvals.
Recommended path