1The Agents-at-Risk ModelThe Agents-at-Risk model turns thousands of agent signals into a ranked list of identity, ownership, and permission failures needing action.Governance and Security 2Policy Templates: Governance by DefaultPolicy templates make approved access the starting point for agents, not a cleanup task after deployment.Governance and Security 3Audit, Compliance and Data ResidencyAudit and discovery prove what agents did, while observability supplies the data and residency rules define where that evidence lives.Governance and Security 4Zero Trust for Non-Human IdentitiesAgents need the same Zero Trust discipline as people: clear identity, least privilege, risk checks, accountable humans, and network controls.Governance and Security 5Windows 365 for AgentsWindows 365 for Agents gives software agents a pooled, stateless, governed place to run without treating them like human users.Governance and Security