Microsoft is expanding Agent 365 with a centralized agent registry that lets administrators view and manage AI agents from Microsoft’s ecosystem and selected third-party platforms. Organizations can synchronize agents from Amazon Bedrock, Google Vertex AI, Salesforce Agentforce, and Databricks through a common governance layer, giving IT admins a single governance layer across multiple AI platforms. The new controls support monitoring, approval workflows, policy enforcement, and lifecycle management to help maintain security and compliance standards as agent deployments grow. Microsoft warns that fragmented management of agents built in different environments, with varying permissions, data access levels, and compliance controls, can increase the risk of security gaps, policy violations, and unchecked AI agent growth, and can create governance gaps similar to shadow IT. Microsoft emphasized that stronger governance introduces additional administrative oversight and management responsibilities. Organizations gain improved visibility and control, but may need to invest time in configuring integrations, establishing approval processes, and maintaining cross-platform governance policies. The centralized approach can reduce risk, but it may also slow the rapid deployment of agents that some IT teams are used to in enterprise environments.