Control 2.9: Defender for Cloud Apps — Copilot Session Controls — Verification & Testing
Test cases and evidence collection for validating session controls.
Test Cases
Test 1: Session Monitoring Activation
- Objective: Confirm Copilot sessions are monitored by Defender for Cloud Apps
- Steps:
- As a test user, sign out of existing Microsoft 365 browser sessions and then reauthenticate to the protected Microsoft 365 web app
- Confirm the session is routed through Conditional Access App Control (lock icon in Edge or
.mcas suffix in another supported browser)
- Navigate to Defender > Cloud Apps > Policies > Policy management and open the relevant session policy report
- Confirm the routed Login activity appears for the test user; if using detailed policies, confirm the supported file or activity events also appear
- Expected Result: Routed browser sessions appear in the policy report; detailed monitored actions appear only when a content-inspecting or block-activities policy is in scope
- Evidence: Policy report entries for the test user, plus screenshots of routed-session indicators
Test 2: Content Inspection Detection
- Objective: Verify content inspection detects sensitive data in supported browser-session actions
- Steps:
- In a supported browser session, attempt a governed file download, upload, or send/share action involving a document that contains test sensitive data
- Verify the session policy content inspection triggers
- Confirm the alert is generated with the correct severity
- Verify the detection details include the sensitive information type
- Expected Result: Content inspection detects and alerts on sensitive data
- Evidence: Alert record with detection details
Test 3: Alert Generation and Delivery
- Objective: Confirm routed-session or anomaly alerts are generated and delivered to the security team
- Steps:
- Trigger a supported session-policy condition (for example, a sensitive file download) or a test anomaly scenario consistent with your tenant policy
- Verify an alert or policy match appears in Defender portal > Alerts or in the relevant policy report
- Confirm email notification is delivered to configured recipients
- Verify alert severity matches the policy configuration
- Expected Result: Alerts generated and delivered within expected timeframe
- Evidence: Alert notification and email confirmation
Test 4: Generative AI App Catalog Coverage
- Objective: Verify the organization has reviewed and governed generative AI app usage via the MDCA catalog
- Steps:
- Navigate to Defender portal > Cloud Apps > Cloud app catalog > filter by Generative AI category
- Confirm the catalog loads and displays generative AI apps
- Open Defender portal > Cloud Apps > Cloud discovery > Discovered apps and identify any generative AI apps used in the organization that are not Microsoft Copilot
- Verify that high-risk discovered generative AI apps have governance actions applied (unsanctioned, blocked through a supported governance stream, or explicitly approved)
- Confirm the sanctioned/unsanctioned review cadence is documented
- Expected Result: Generative AI catalog and Cloud Discovery have both been reviewed; high-risk apps are governed; approval or prohibition decisions are documented
- Evidence: App catalog screenshot, discovered-app evidence, and governance policy/configuration for high-risk apps
Test 5: Agent Threat Detection Verification
- Objective: Confirm agent threat detection is operational for Copilot agent deployments
- Steps:
- Confirm the tenant holds a Microsoft Agent 365-eligible license using the Control 2.9 qualifying-license guidance; Microsoft 365 E5 is not an exclusive prerequisite. This is required as of July 1, 2026 for Copilot Studio/Foundry agent threat-detection coverage
- Navigate to Defender portal > Assets > AI agents and confirm Copilot agent deployments appear with a current risk level
- Navigate to Defender portal > Incidents & alerts and filter for agent-related alerts
- Review any existing agent-related incidents to verify the detection is surfacing actionable intelligence
- Verify at least one custom agent anomaly detection rule is configured
- Confirm agent threat alert routing: alerts should reach the SOC or security team within the defined SLA
- Expected Result: The tenant's Agent 365 license is confirmed, agents appear in the AI agent inventory with a risk level, alerts are routing correctly, and custom rules are configured
- Evidence: Agent 365 license confirmation; agent inventory screenshot; sample agent incident records; alert routing confirmation
Evidence Collection
| Evidence Item |
Format |
Storage Location |
Retention |
| Session policy configuration |
Screenshot/PDF |
Compliance evidence repository |
7 years |
| Policy report / activity log samples |
CSV |
Compliance evidence repository |
7 years |
| Alert records |
CSV |
Compliance evidence repository |
7 years |
| Content inspection test results |
PDF |
Compliance evidence repository |
7 years |
Compliance Mapping
| Regulation |
Requirement |
How This Control Supports It |
| FINRA Rule 3110 |
Supervisory system monitoring |
Session controls support compliance with AI interaction monitoring requirements |
| SEC Rule 17a-4 |
Electronic communication monitoring |
Session logging helps meet communication monitoring obligations |
| FFIEC Handbook |
Security monitoring |
Real-time session controls support compliance with security monitoring requirements |
| NIST CSF |
DE.CM-1 Network monitoring |
Session controls provide monitoring for AI workloads |
| - Back to Control 2.9 |
|
|