KQL Search

Query Details

M365 Copilot Extensions Threat Monitoring

Query

// M365 Copilot Extensions Threat Monitoring

// https://www.linkedin.com/posts/0x534c_cybersecurity-generativeai-m365-activity-7250789113249837056-D6rZ/
// This just keeps getting better and better! 😂 I absolutely refuse to accept Message Center MC908119, which claims that ‘Admins will lose tenant-level control over who can use Copilot agents.’ 😤 As a countermeasure, I’ve developed a Sentinel analytics rule that monitors all external URLs accessed by M365 Copilot extensions against my Threat Intelligence database for potential malicious activities. This rule is linked with a logic playbook automation to mark the user as compromised and isolate their access. 💯😘

// ConfusedPilot Attack Vector:
// https://www.linkedin.com/posts/0x534c_confusedpilot-activity-7252113628470943744-BgxN/

// M365 Copilot Extensions Threat Monitoring (Sentinel)
// https://www.linkedin.com/posts/0x534c_cybersecurity-generativeai-copilot-activity-7251091656249090048-GfSg/

CloudAppEvents
| where TimeGenerated > ago(1h)
| where ActionType == @"CopilotInteraction"
| extend UserID = tostring(RawEventData.UserId)
| extend CopilotData = todynamic(RawEventData.CopilotEventData)
| extend CopilotPlugin = tostring(CopilotData.AISystemPlugin[0].Id)
| where isnotempty(CopilotPlugin)
| extend PluginAccessURL = tostring(CopilotData.AccessedResources)
| mv-expand todynamic(PluginAccessURL)
| where PluginAccessURL has "SiteUrl"
| extend Url = tostring(PluginAccessURL.SiteUrl)
| extend Domain = tostring(parse_url(Url).Host)
| extend Action = tostring(PluginAccessURL.Action)
| join ThreatIntelligenceIndicator on $left.Domain == $right.DomainName


// MITRE ATT&CK
// T1116 Browser Extensions

Explanation

This query is designed to monitor potential threats associated with Microsoft 365 Copilot extensions by analyzing cloud application events. Here's a breakdown of what the query does in simple terms:

  1. Time Filter: It looks at events generated in the last hour.

  2. Action Type: It specifically focuses on events where the action type is "CopilotInteraction," indicating interactions with Copilot extensions.

  3. User and Copilot Data Extraction:

    • It extracts the user ID and detailed Copilot event data from the raw event data.
    • It identifies the specific Copilot plugin involved in the interaction.

  4. URL and Domain Extraction:

    • It extracts URLs accessed by the Copilot plugin.
    • It expands these URLs to find specific site URLs accessed.
    • It parses these URLs to extract the domain name.

  5. Threat Intelligence Matching:

    • It compares the extracted domain names against a threat intelligence database to identify potential malicious activities.

  6. Automation and Response:

    • If a match is found, the rule is linked with an automated playbook that marks the user as compromised and isolates their access to prevent further threats.

  7. Reference to MITRE ATT&CK:

    • The query references the MITRE ATT&CK framework, specifically T1116, which relates to the use of browser extensions for malicious purposes.

Overall, this query is part of a security monitoring strategy to detect and respond to potential threats from external URLs accessed by M365 Copilot extensions, ensuring that any suspicious activity is promptly addressed.