Microsoft To Let IT Admins Monitor Risky AI Prompts in Purview

Microsoft is expanding its enterprise AI governance tools with new monitoring capabilities inside Microsoft Purview Insider Risk Management. The update arrives shortly after the company gave IT administrators more control over removing Microsoft Copilot from managed environments.

Announced through Message Center update MC1304292, the new feature will let authorized security and IT personnel review potentially risky interactions with AI services. Microsoft plans to begin preview rollout later this month, while general availability is expected next month.

Microsoft wants tighter AI oversight

The upcoming capability allows investigators to view prompts employees send to AI platforms alongside AI-generated responses connected to flagged activity. In some cases, both prompts and responses may appear in plaintext during investigations.

Microsoft says the feature targets growing concerns around “shadow AI” usage inside organizations. Many enterprises now worry about employees accidentally sharing sensitive data, internal documents, or proprietary code with external AI tools.

The company positions the feature as part of a broader push toward AI governance and enterprise compliance. Organizations will be able to detect risky usage patterns, potential intellectual property theft, policy violations, and accidental data exposure tied to AI interactions.

Privacy protections remain in place

Microsoft also emphasized privacy protections within the system. Users can remain pseudonymized during initial reviews, while investigators must have proper authorization before deanonymizing individuals during active investigations.

According to Microsoft, role-based access controls and audit logging remain active to help balance enterprise security requirements with employee privacy protections. The company says every access event continues to generate audit records for compliance tracking.

Part of Microsoft’s wider AI governance strategy

The new monitoring tools also align with Microsoft’s wider efforts around AI oversight across enterprise products, including recent initiatives connected to Microsoft Edge and Purview designed to limit unmanaged AI usage inside corporate environments.

In other Microsoft news, the company is reportedly removing the Edge sidebar as it shifts focus toward Copilot experiences. Microsoft is also preparing mobile apps for Copilot Cowork.

Via Neowin