Detection rules › YARA-L
O365 Login Activity To Azure AD PowerShell App
Logins to Azure AD PowerShell app can have legitimate purposes, but are also abused to gain access to user information. Programmatic access to Entra ID (Azure AD) should generally be through apps, so reviewing these activities is needed.
MITRE ATT&CK coverage
| Tactic | Techniques |
|---|---|
| Initial Access |
Telemetry coverage
| Platform | Record / event type |
|---|---|
| Microsoft 365 | UserLoggedIn |
Rules detecting the same action
These rules filter on the same operation.
- Azure Login Bypassing Conditional Access Policies (Sigma)
- Hunt for suspicious sign-in to Entra ID Using Extrnal Call Method (YARA-L)
- M365 Identity OAuth Flow by First-Party Microsoft App from Multiple IPs (Elastic)
- M365 Identity OAuth Flow by User Sign-in to Device Registration (Elastic)
- M365 Identity OAuth Phishing via First-Party Microsoft Application (Elastic)
- M365 Identity OAuth ROPC Grant via Legacy Authentication Client (Elastic)
- M365 Potential AiTM UserLoggedIn via Office App (Tycoon2FA) (Elastic)
- O365 AD PowerShell App Login Subsequent Activity (YARA-L)
Rule body
// Copyright 2025 Google LLC. Licensed under Apache-2.0.
rule o365_login_activity_to_azure_ad_powershell_app {
meta:
author = "Google Cloud Security"
description = "Logins to Azure AD PowerShell app can have legitimate purposes, but are also abused to gain access to user information. Programmatic access to Entra ID (Azure AD) should generally be through apps, so reviewing these activities is needed."
rule_id = "mr_0cc56eaf-4176-444d-874d-9780bbb29d4f"
rule_name = "O365 Login Activity To Azure AD PowerShell App"
tactic = "TA0001"
technique = "T1078.004"
type = "hunt"
platform = "azure"
data_source = "o365"
severity = "Medium"
priority = "Medium"
events:
$login.metadata.event_type = "USER_LOGIN"
$login.metadata.product_event_type = "UserLoggedIn"
$login.metadata.product_name = "Office 365"
$login.metadata.vendor_name = "Microsoft"
$login.target.resource.product_object_id = "1b730954-1685-4b74-9bfd-dac224a7b894"
$login.security_result.action = "ALLOW"
// Below filters ADFS Sync Account that runs every 30 minutes between AD and Entra ID (for federated environments); can use display name or email address
//$login.target.user.userid != "insert your email address here" and
//$login.target.user.user_display_name != "On-Premises Directory Synchronization Service Account"
$login.target.user.userid = $userid
match:
$userid over 5m
outcome:
$risk_score = 65
$event_count = count_distinct($login.metadata.id)
$security_summary = array_distinct($login.security_result.summary)
$user_agent = array_distinct($login.network.http.user_agent)
$country_region_login_attempt = array_distinct($login.principal.ip_geo_artifact.location.country_or_region)
//added to populate alert graph with additional context
$principal_ip = array_distinct($login.principal.ip)
$target_user_userid = array_distinct($login.target.user.userid)
condition:
$login
}
YARA-L rule structure
Condition
Fires when at least one $login event in the 5m window.
Events
$login
metadata.product_event_type = "UserLoggedIn"target.resource.product_object_id = "1b730954-1685-4b74-9bfd-dac224a7b894"security_result.action = "ALLOW"
Match and correlation
Outcome
Outcome variables add context to a detection. When the rule generates an alert, $risk_score is displayed with it.
| Field | Expression |
|---|---|
risk_score | 65 |
event_count | count_distinct($login.metadata.id) |
security_summary | array_distinct($login.security_result.summary) |
user_agent | array_distinct($login.network.http.user_agent) |
country_region_login_attempt | array_distinct($login.principal.ip_geo_artifact.location.country_or_region) |
principal_ip | array_distinct($login.principal.ip) |
target_user_userid | array_distinct($login.target.user.userid) |