Microsoft Teams is introducing support for third-party synthetic audio and video detection, allowing organizations to integrate certified deepfake detection solutions directly into the meeting experience. When a detection solution identifies signs of manipulated or AI-generated audio or video, Teams surfaces warnings and risk indicators to meeting participants — helping users recognize potential impersonation attempts before acting on what they see or hear. For organizations where meeting security and identity trust are critical, this update adds a meaningful layer of protection against an increasingly accessible threat vector. The feature is currently in development and expected to begin rolling out in November 2026.

Microsoft Teams is introducing a single pane in the Teams Admin Center for monitoring and troubleshooting meeting and call health across the organization. Administrators can identify recurring patterns, investigate issues affecting users, and proactively address problems in both active and completed meetings and calls — without navigating across multiple tools or reports. The feature is currently rolling out for Commercial and GCC environments.

Microsoft Purview Insider Risk Management is extending risk detection to Microsoft Fabric by introducing ready-to-use risk indicators based on Power BI activity. Organizations can now incorporate these indicators into data theft and data leaks policies, capturing risky behavior in Power BI — such as unusual data exports or access patterns — alongside signals from other Microsoft 365 workloads. For compliance teams managing insider risk across an expanding data estate, this update closes a gap where business intelligence activity was previously outside IRM’s detection scope. The feature is currently in development and expected to begin rolling out in October 2026.

Microsoft Purview is introducing Classifier Simulation Mode, enabling admins to test and optimize custom classifiers against production data before publishing them to live environments. Custom classifiers with overly broad logic or inefficient patterns can generate excessive false positives, high match volumes, and scanning latency — Simulation Mode surfaces these issues before they affect enforcement. For compliance teams managing sensitive information types at scale, this update removes the guesswork from classifier deployment and replaces it with a validated, performance-tested path to production. The feature is currently in development and expected to begin rolling out in November 2026.

Microsoft Purview Data Loss Prevention and Information Protection auto-labeling are expanding beyond Microsoft services to cover non-Microsoft connected apps including Google Workspace, Box, Dropbox, Salesforce, ServiceNow, Amazon Web Services, and Cisco Webex. Organizations can now apply the same classifiers, sensitive information types, and sensitivity labels used across Microsoft 365 to detect and protect sensitive data shared through these platforms — without managing separate DLP tooling for each service. For compliance teams managing data across a mixed cloud environment, this update brings a meaningful portion of the shadow IT and cross-platform data exposure risk under centralized Purview governance. The feature is currently in development and expected to begin rolling out in September 2026.

Microsoft Purview Data Loss Prevention is expanding to calendar events in Outlook for Mac and mobile, enabling organizations to detect and prevent sensitive information from being shared through meeting subjects, descriptions, and invite content. When a policy is triggered, users receive policy tips, remediation guidance, or enforcement actions — the same protection experience already available for email, now applied to a surface that has historically operated outside DLP coverage. For organizations managing compliance across all communication channels, this update closes a meaningful gap where sensitive data could leave through calendar invites without any policy enforcement in place. The feature is currently in development and expected to begin rolling out in September 2026.

Microsoft Teams is introducing app-centric management in the Teams Admin Center for GCC High, giving admins granular control over which users, groups, or the entire tenant can install individual Teams apps. Admins can set a default behavior for newly published apps and override availability per app — choosing between all users, specific users and groups, or no one — replacing the broader app permission policy model with per-app precision. Existing app permission policies are migrated automatically to preserve current availability. The feature is currently in development and expected to begin rolling out in September 2026.

Microsoft OneDrive is adding the ability for users to exclude specific folders from syncing on Windows and Mac, keeping local-only content such as PowerShell modules, Visual Studio configuration, and node modules off OneDrive entirely. Administrators can define folder exclusion rules at the organizational level, giving IT teams consistent control over what does and doesn’t sync across managed devices. For organizations managing developer workstations or devices with large local tooling directories, this update reduces unnecessary storage consumption and keeps cloud sync focused on content that belongs there. The feature is currently rolling out.

Microsoft Teams now allows administrators to enable automatic recording and transcription for Teams call queues, ensuring calls answered by representatives are captured without requiring manual action from agents. Admins can configure the feature per queue through the Teams admin center or PowerShell, with granular controls over recording, transcription, and agent access to recordings. Recordings are stored in SharePoint and accessible through the Queues app call history, giving organizations a consistent, auditable record of customer interactions. The feature is currently rolling out for Commercial and GCC environments and requires a Teams Premium license for the Queues app.

Microsoft Purview sensitivity labels now apply directly to Microsoft Entra cloud security groups, closing a governance gap that left security groups outside the classification model you already use everywhere else. The same labels you publish in the Purview portal and configure for Microsoft 365 groups and sites now carry over automatically, with no separate label setup required. That means the access controls and classification tied to sensitive data extend to the security groups actually gating that access, not just the collaboration groups around it. It’s a small config change with an outsized governance payoff for any org running a mature labeling strategy. The feature is currently in development and expected to begin rolling out in August 2026.