Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Important
You need to be part of the Frontier preview program to get early access to Microsoft Agent 365. Frontier connects you directly with Microsoft’s latest AI innovations. Frontier previews are subject to the existing preview terms of your customer agreements. As these features are still in development, their availability and capabilities may change over time.
Learn how to protect your data with Microsoft Purview.
Important
This capability is rolling out in phases and the rollout is expected to complete by end of December.
Secure agents by default with Microsoft Purview
Depending on the type of agent, Microsoft Purview supports secure by default capabilities including:
- Audit trails that log agent activities.
- Data security controls to detect sensitive data in agent interactions.
- Compliance monitoring to evaluate agents for compliance gaps and identify areas needing attention.
Based on the secure by default capabilities, IT admins inside Microsoft 365 admin center can understand the agent’s aggregated activities, sensitive data interactions, and AI baseline assessment. These insights can be expanded upon engagement with the data security admin through the Microsoft Purview portal.
AI observability and data security in Purview
Inside Microsoft Purview, data security admins can:
- See a list of their agent inventory.
- Use Purview signals to prioritize investigation and remediation of agents.
- Drill into agent context, activities with sensitive date, and remediation recommendations.
- View all agent activity in audit logs.
View agent observability logs in Purview
Sign in to Purview with your admin credentials.
Go to View All Solutions and select Audit.
In the search user interface, locate the Record Types field. Select the following record types:
- AIExecuteTool
- AIInvokeAgent
- AIInferenceCall
Select Search. The search results are displayed in a list with their status.
Select a search result to view detailed events, including:
- Agent invocation logs
- Tool execution details
- Inference call traces