Detection rules › Elastic
Entra ID Register Device with Unusual User Agent (Azure AD Join)
Detects successful Microsoft Entra ID audit events for Register device where additional details indicate an Azure AD join and the recorded user agent is not one of the common native registration clients (Dsreg, DeviceRegistrationClient, or Dalvik-based Android enrollment). Legitimate Windows and standard mobile enrollment flows often present predictable user-agent strings; unexpected clients may reflect scripted registration, third-party tooling, or adversary-driven device registration used for persistence or token abuse. Baseline approved provisioning tools and MDM integrations before tuning.
MITRE ATT&CK coverage
| Tactic | Techniques |
|---|---|
| Persistence | T1098.005 Account Manipulation: Device Registration |
Event coverage
| Provider | Event |
|---|---|
| Entra-AuditLogs | Register device |
Rule body elastic
[metadata]
creation_date = "2026/05/15"
integration = ["azure"]
maturity = "production"
updated_date = "2026/05/15"
[rule]
author = ["Elastic"]
description = """
Detects successful Microsoft Entra ID audit events for Register device where additional details indicate an Azure AD join
and the recorded user agent is not one of the common native registration clients (Dsreg, DeviceRegistrationClient, or
Dalvik-based Android enrollment). Legitimate Windows and standard mobile enrollment flows often present predictable
user-agent strings; unexpected clients may reflect scripted registration, third-party tooling, or adversary-driven device
registration used for persistence or token abuse. Baseline approved provisioning tools and MDM integrations before tuning.
"""
false_positives = [
"""
Custom device management agents, OEM enrollment clients, or updated Microsoft clients that use new user-agent strings
may match. Add exclusions for known `azure.auditlogs.properties.userAgent` values or enrollment programs after review.
""",
]
from = "now-9m"
index = ["logs-azure.auditlogs-*"]
language = "kuery"
license = "Elastic License v2"
name = "Entra ID Register Device with Unusual User Agent (Azure AD Join)"
note = """## Triage and analysis
### Investigating Entra ID Register Device with Unusual User Agent (Azure AD Join)
Review `azure.auditlogs.properties.initiated_by.user.userPrincipalName`, source IP fields on the audit event,
`azure.auditlogs.properties.target_resources.0.display_name` for the device name, and
`azure.correlation_id` for related audit entries.
Compare `azure.auditlogs.properties.userAgent` to your organization's standard Autopilot, Intune, and Windows enrollment
clients.
### Possible investigation steps
- Confirm whether the user intentionally joined a device and whether the user agent matches a known provisioning package.
- Pivot to `azure.signinlogs` for the same principal and timeframe for risky sign-ins or token broker activity.
- Search for other `Register device` events from the same IP or user agent across the tenant.
### Response and remediation
- If malicious, remove the device in Entra ID, revoke refresh and primary refresh tokens for the user, and reset credentials per policy.
- Tighten device registration and join controls via Conditional Access and device compliance policies.
"""
references = [
"https://learn.microsoft.com/en-us/entra/identity/devices/concept-directory-join",
"https://github.com/dirkjanm/ROADtools",
]
risk_score = 47
rule_id = "d4695889-0410-4e7b-a4aa-59be525a11a6"
severity = "medium"
tags = [
"Domain: Cloud",
"Domain: Identity",
"Data Source: Azure",
"Data Source: Microsoft Entra ID",
"Data Source: Microsoft Entra ID Audit Logs",
"Use Case: Threat Detection",
"Tactic: Persistence",
"Resources: Investigation Guide",
]
timestamp_override = "event.ingested"
type = "query"
query = '''
data_stream.dataset:"azure.auditlogs" and event.action:"Register device" and
event.outcome:(success or Success) and
azure.auditlogs.properties.userAgent:(* and not (Dsreg* or DeviceRegistrationClient or Dalvik*)) and
azure.auditlogs.properties.additional_details.value:"Azure AD join"
'''
[rule.investigation_fields]
field_names = [
"@timestamp",
"event.action",
"event.outcome",
"azure.auditlogs.properties.userAgent",
"azure.auditlogs.properties.additional_details.value",
"azure.auditlogs.properties.initiated_by.user.userPrincipalName",
"azure.auditlogs.properties.initiated_by.user.ipAddress",
"azure.auditlogs.properties.target_resources.0.display_name",
"azure.auditlogs.properties.target_resources.0.id",
"azure.correlation_id",
]
[[rule.threat]]
framework = "MITRE ATT&CK"
[[rule.threat.technique]]
id = "T1098"
name = "Account Manipulation"
reference = "https://attack.mitre.org/techniques/T1098/"
[[rule.threat.technique.subtechnique]]
id = "T1098.005"
name = "Device Registration"
reference = "https://attack.mitre.org/techniques/T1098/005/"
[rule.threat.tactic]
id = "TA0003"
name = "Persistence"
reference = "https://attack.mitre.org/tactics/TA0003/"
Stages and Predicates
Stage 1: query
data_stream.dataset:"azure.auditlogs" and event.action:"Register device" and
event.outcome:(success or Success) and
azure.auditlogs.properties.userAgent:(* and not (Dsreg* or DeviceRegistrationClient or Dalvik*)) and
azure.auditlogs.properties.additional_details.value:"Azure AD join"
Exclusions
Top-level NOT(...) conjuncts: predicates this rule actively suppresses.
| Field | Kind | Excluded values |
|---|---|---|
azure.auditlogs.properties.userAgent | eq | DeviceRegistrationClient |
azure.auditlogs.properties.userAgent | starts_with | Dalvik |
azure.auditlogs.properties.userAgent | starts_with | Dsreg |
Indicators
Each row is a field, operator, and value that the rule matches. The corpus column counts how many other rules in the catalog look for the same combination: high numbers point to widely-used, community-vetted indicators. Blank or 1 shows that the indicator is specific to this rule.
| Field | Kind | Values |
|---|---|---|
azure.auditlogs.properties.additional_details.value | eq |
|
azure.auditlogs.properties.userAgent | is_not_null | |
data_stream.dataset | eq |
|
event.action | eq |
|
event.outcome | in |
|