← Back to Cybersecurity News Center
Severity
HIGH
CVSS
7.5
Priority
0.368
Executive Summary
Threat actors are injecting malicious content into legitimate Microsoft Azure Monitor alert emails sent from azure-noreply@microsoft.com, causing them to pass all standard email authentication checks (SPF, DKIM, DMARC). Recipients receive what appears to be an official Microsoft Security Team message containing a fabricated $389.90 Windows Defender billing charge and instructions to call an attacker-controlled phone number. Any organization using Azure Monitor alert notifications is potentially exposed; the primary business risks are credential theft, unauthorized network access, and financial fraud via social engineering.
Technical Analysis
Attack vector: Threat actors create or modify Azure Monitor alert rules, inserting arbitrary content into the alert description field.
When Azure Monitor fires the alert, Microsoft's own notification pipeline delivers the crafted message from azure-noreply@microsoft.com with valid SPF, DKIM, and DMARC authentication, bypassing email security gateways and anti-phishing controls that rely on sender authentication.
The lure impersonates the Microsoft Security Team and presents a fabricated billing charge to coerce recipients into calling attacker-controlled numbers (callback phishing / TOAD, Telephone-Oriented Attack Delivery).
Over the phone, attackers likely pursue credential harvesting, remote access tool installation (T1219), or direct financial fraud. No CVE is assigned; the behavior exploits a legitimate platform feature, not a vulnerability in the traditional sense. Relevant CWEs: CWE-284 (Improper Access Control, alert field accepts arbitrary content), CWE-20 (Improper Input Validation, no sanitization of alert description content before inclusion in outbound email), CWE-940 (Improper Verification of Source of a Communication Channel, recipients cannot distinguish legitimate from weaponized alert emails by sender authentication alone). MITRE ATT&CK coverage: T1566 (Phishing), T1566.004 (Spearphishing Voice), T1036 / T1036.005 (Masquerading), T1598 (Phishing for Information), T1204.002 (User Execution: Malicious File, applicable if phone call leads to remote tool delivery), T1219 (Remote Access Software), T1078 (Valid Accounts), T1059 (Command and Scripting Interpreter, post-access), T1199 (Trusted Relationship, abuse of Microsoft's trusted sending infrastructure). No patch is currently available; Microsoft has not issued a platform-level fix as of reporting. Mitigation is defensive and procedural.
Action Checklist IR ENRICHED
Triage Priority:
IMMEDIATE
Escalate to IR management and notify CISO if more than 5 users report receiving the phishing email, if any user reports calling the attacker number, or if unauthorized alert rules are confirmed in production subscriptions; consider external IR firm engagement if alert rules are discovered on critical infrastructure or healthcare subscriptions.
Step 1, Immediate: Alert all Azure administrators and help desk staff that azure-noreply@microsoft.com alert emails may contain attacker-injected content; instruct staff not to call any phone number presented in an Azure Monitor alert email without out-of-band verification.
Preparation
NIST 800-61r3 §2.1 (organization preparation) and §2.3 (tools and resources)
NIST IR-1 (incident response policy)
NIST AT-2 (security awareness training)
CIS 17.1 (communication templates for security incidents)
Compensating Control
Create and distribute a one-page alert via email, Slack, or internal wiki with a screenshot of the lure and the rule: 'Do not call any phone number in an Azure Monitor alert email. Instead, verify the alert by logging into your Azure portal directly and checking the alert rule.' Document distribution in a shared spreadsheet with timestamps and recipient confirmation.
Preserve Evidence
Capture the original phishing email headers (Message-ID, X-Originating-IP, X-Mailer, Received chain) before forwarding to security team. Preserve the alert email in a dedicated folder for later forensic analysis. Document the timestamp when staff were notified to establish timeline of awareness.
Step 2, Immediate: Review Azure Monitor alert rules across all subscriptions for unauthorized or unfamiliar alert configurations, paying particular attention to alert descriptions containing phone numbers, billing language, or impersonation of Microsoft support teams.
Detection & Analysis
NIST 800-61r3 §3.2.2 (initial analysis and hypothesis) and §3.2.3 (event documentation)
NIST CA-7 (continuous monitoring)
NIST SI-4 (information system monitoring)
CIS 8.2 (configuration baseline and integrity checking)
Compensating Control
Export all Azure Monitor alert rules via Azure PowerShell: `Get-AzMetricAlertRuleV2 -ResourceGroupName '*' | Select-Object Name, Description, CreatedTime, LastModifiedTime | Export-Csv alerts.csv`. Manually review the Description field for keywords: phone numbers (regex: \d{3}-\d{3}-\d{4}), 'billing', 'charge', 'Defender', 'support team', 'confirm', 'verify payment'. Cross-reference alert creators against known administrators in a spreadsheet.
Preserve Evidence
Capture Azure Activity Log entries for resource type 'microsoft.insights/alertrules' with operations 'CreateOrUpdateMetricAlertRule' and 'DeleteMetricAlertRule' from the past 90 days. Export via: `Get-AzActivityLog -ResourceGroupName '*' -ResourceType 'microsoft.insights/alertrules' -StartTime (Get-Date).AddDays(-90)`. Document the exact alert rule ID, description text, and creator identity for each suspicious rule.
Step 3, Detection: Query Microsoft 365 mail flow logs and email security platform logs for inbound messages from azure-noreply@microsoft.com containing keywords: 'Windows Defender', 'billing', 'charge', '$', 'call', '1-8', or phone number patterns, flag any matches for analyst review.
Detection & Analysis
NIST 800-61r3 §3.2.1 (detection and analysis overview) and §3.2.4 (event correlation)
NIST SI-4(1) (system monitoring with organization-defined monitoring tools)
NIST AU-6(9) (audit record review with data analytics)
CIS 8.8 (e-mail and web browser protections)
Compensating Control
Query Exchange Online Mail Flow logs using: `$logs = Get-MessageTrace -SenderAddress 'azure-noreply@microsoft.com' -StartDate (Get-Date).AddDays(-30) -EndDate (Get-Date); $logs | Select-Object * | Export-Csv azure_alert_emails.csv`. Use a text editor or grep to search the exported body/subject for patterns: `grep -iE '(Windows Defender|billing|charge|\$|call|1-8[0-9]{2})' azure_alert_emails.csv`. Document all matches with recipient, timestamp, and subject line.
Preserve Evidence
Preserve the full email headers (Internet Message ID, received timestamps, DKIM/SPF/DMARC results) and message body for each flagged email. Export message headers using: `Get-Message -Identity '<MessageID>' | Export-CliXml message_headers.xml`. If email security platform is present, export threat logs with sender, recipient, keywords detected, and timestamp. Create a list of all recipient addresses and forward to notification step.
Step 4, Assessment: Audit Azure Monitor alert rule creation and modification logs (Azure Activity Log, resource type: microsoft.insights/alertrules) for changes made by unfamiliar identities or service principals over the past 90 days; cross-reference with Azure AD sign-in logs for anomalous access.
Detection & Analysis
NIST 800-61r3 §3.2.5 (impact analysis) and §3.4.1 (evidence gathering and handling)
NIST AU-2 (audit events)
NIST AU-12 (audit generation)
NIST IA-4 (identifier management)
CIS 6.2 (managed logging and log alerting)
Compensating Control
Export Azure Activity Log: `Get-AzActivityLog -ResourceType 'microsoft.insights/alertrules' -StartTime (Get-Date).AddDays(-90) | Select-Object EventTimestamp, Caller, OperationName, ResourceId, StatusCode, Properties | Export-Csv activity_log.csv`. Export Azure AD sign-in logs: `Export-AzureADSignInLog -StartDate (Get-Date).AddDays(-90) -EndDate (Get-Date) | Export-Csv signin_log.csv`. Cross-reference 'Caller' identity from activity log against a list of known administrators and approved service principals. Flag any entries where Caller is unfamiliar or sign-in shows unusual time, location, or IP address.
Preserve Evidence
Preserve the full activity log entry including EventTimestamp, Caller (user principal name and object ID), OperationName, Resource ID, and Properties JSON. For each flagged Caller, capture Azure AD sign-in logs showing all login attempts from that identity during the 90-day window, including IP address, device, browser, and MFA status. Document any gaps in audit logs (retention limits, deletions).
Step 5, Communication: Notify end users and executives via a verified internal channel that this campaign is active; provide examples of the lure message and clear instructions for reporting suspected callback phishing attempts to the security team.
Containment
NIST 800-61r3 §2.3.4 (providing user awareness training) and §3.3.1 (containment planning)
NIST AT-2 (security awareness and training)
NIST IR-4 (incident handling)
CIS 17.8 (incident response process)
Compensating Control
Draft a templated security alert email with: threat name, a sanitized screenshot of the lure (phone number redacted), key indicators (billing language, request to call), and a reporting mechanism (email to security-team@company.com or phone number for verified security desk). Distribute via email from the CIO/Security Officer with confirmation of read receipts. Post the same alert in a pinned message in Slack #general and on the internal security wiki. Log distribution timestamp and recipient count.
Preserve Evidence
Capture and archive the original phishing email and any variant samples discovered during Step 3. Document the exact lure text, phone number(s), and alert rule descriptions. Record the timestamp and audience of each notification sent. If end users report similar emails, capture those reports with headers and timestamps to assess campaign scope and evolution.
Step 6, Long-term: Evaluate Azure Monitor alert notification configurations to restrict who can create or modify alert rules with broad distribution; consider implementing Azure Policy to enforce alert description content standards or limit external email targets for alert notifications.
Eradication
NIST 800-61r3 §3.4.1 (eradication) and §3.4.2 (eradication strategy)
NIST AC-2 (account management)
NIST AC-3 (access enforcement)
NIST CA-9 (internal system connections)
CIS 5.3 (access control list management)
Compensating Control
Manually audit Azure Monitor alert rule creators using Azure portal: IAM > Roles > Custom roles > filter by operations containing 'alertrules'. Create a custom Azure role that denies 'microsoft.insights/alertrules/write' except for a named security team group. Assign this role at the subscription level. For alert descriptions, create a PowerShell script to scan all alert rules monthly and flag any containing regex patterns (phone numbers, billing keywords, 'call'). Store the script in Azure Automation Runbook and log findings to a CSV for manual review. Document the control in a runbook wiki page.
Preserve Evidence
Capture the current RBAC assignments for Azure Monitor Alert rule creator permissions. Document the custom role definition (JSON) and assignment scope. Record the baseline of all alert rules before and after the control is applied, including rule ID, creator, description, and modification history. Establish a baseline for future compliance audits.
Step 7, Long-term: Update security awareness training to include TOAD / callback phishing as a distinct threat pattern, with specific guidance on Microsoft billing communications (Microsoft does not issue billing charges via Azure Monitor alert emails).
Post-Incident
NIST 800-61r3 §3.5.1 (post-incident activities) and NIST SP 800-50 (security awareness and training)
NIST AT-2 (security awareness and training)
NIST AT-3 (role-based security training)
CIS 17.7 (security awareness training program)
Compensating Control
Add a 5-minute module to annual security training covering: what TOAD (Toll Fraud and Authenticated Callback Phishing) is, how this campaign weaponizes legitimate Azure Monitor emails, the red flags (billing language, phone numbers, urgency), and the correct response (verify in Azure portal, do not call). Distribute via LMS or video link. Create a one-page quick reference card for help desk and post to the security intranet. Track completion rates and quiz scores. Update the company security policy to state: 'Microsoft never sends billing charges via alert emails. Always verify security alerts by logging into the Microsoft portal directly.'
Preserve Evidence
Document training module creation date, audience, completion metrics, and assessment scores. Capture a sample of the training material and quick reference card. Track any phishing simulation results related to this threat pattern before and after training deployment to measure awareness improvement.
Recovery Guidance
After unauthorized alert rules are removed and email filters are deployed, conduct a full review of Azure Monitor alert rule backups and compare against the current state to identify any rules created during the attack window that were subsequently deleted. Audit all Azure services that triggered alerts during the 90-day lookback period to ensure alert content was not maliciously modified. Provide affected users with verification guidance: direct them to log into the Azure portal, navigate to Monitor > Alerts, and compare alert rule descriptions against any emails they received to confirm legitimacy.
Key Forensic Artifacts
Azure Activity Log entries for resource type 'microsoft.insights/alertrules' (90-day lookback): operations CreateOrUpdateMetricAlertRule, DeleteMetricAlertRule, and ModifyAlertRule with Caller identity and timestamp
Azure AD sign-in logs for all identities that created or modified alert rules, including IP address, user agent, device compliance, and MFA status
Exchange Online mail flow logs and message headers for all inbound messages from azure-noreply@microsoft.com (30-day lookback minimum) with body content containing keywords: 'billing', 'charge', 'Defender', 'call', phone number patterns
Azure Monitor alert rule export (JSON or PowerShell Get-AzMetricAlertRuleV2 output) showing rule name, description, creator, creation/modification timestamps, and notification targets for all subscriptions
Email security platform threat logs (if available) showing sender reputation, DKIM/SPF/DMARC pass/fail status, URL rewrite events, and phishing confidence scores for messages from azure-noreply@microsoft.com
Detection Guidance
Primary detection surface is Azure Activity Log and Microsoft 365 mail flow. Key indicators and queries:
- Azure Activity Log, filter for operationName 'microsoft.insights/alertrules/write' or 'microsoft.insights/scheduledqueryrules/write' by identities outside your expected administrator set; look for creation or modification events from unfamiliar service principals or guest accounts.
- Microsoft 365 / Exchange mail flow logs, search for messages where SenderAddress = 'azure-noreply@microsoft.com' AND body or subject contains phone number patterns (e.g., regex: \b(1[-.]?\(?8[0-9]{2}\)?[-.]?[0-9]{3}[-.]?[0-9]{4})\b) or billing terms ('charge', 'invoice', 'Windows Defender', '$').
- Defender for Office 365 or equivalent, build a custom detection rule triggering on authenticated mail from azure-noreply@microsoft.com where body content matches social engineering keyword clusters.
- Endpoint telemetry, if a user has already called a number and interacted with an attacker, look for: new remote access tool installations (AnyDesk, TeamViewer, ScreenConnect), LOLBin execution chains (mshta, wscript, cscript, powershell) spawned from browser or email client processes, and new scheduled tasks or services created within 30 minutes of an anomalous phone call. Behavioral indicator: any user-reported phone call to a 'Microsoft billing' number received via an Azure email alert is a confirmed TOAD attempt and warrants immediate endpoint investigation.
Indicators of Compromise (3)
| Type | Value | Context | Confidence |
| DOMAIN |
azure-noreply@microsoft.com (sending address) |
Legitimate Microsoft sending address weaponized by injecting attacker content into Azure Monitor alert descriptions; presence alone is not malicious — inspect email body for phone numbers and billing lures |
high |
| URL |
No attacker-controlled URLs reported in available sources |
Campaign relies on phone callback rather than embedded URLs; standard URL-based IOCs are not applicable for this vector |
high |
| PATTERN |
Fabricated charge: $389.90 Windows Defender billing |
Lure amount reported in BleepingComputer coverage; may vary across campaign variants — treat any Azure alert email referencing Defender billing as suspicious regardless of amount |
medium |
Compliance Framework Mappings
T1199
T1566
T1204.002
T1036
T1219
T1566.004
+4
AT-2
CA-7
SC-7
SI-3
SI-4
SI-8
+8
6.1
6.2
16.10
6.3
14.2
8.2
164.312(a)(1)
164.312(d)
164.308(a)(5)(i)
MITRE ATT&CK Mapping
T1199
Trusted Relationship
initial-access
T1566
Phishing
initial-access
T1036
Masquerading
defense-evasion
T1219
Remote Access Tools
command-and-control
T1078
Valid Accounts
defense-evasion
T1598
Phishing for Information
reconnaissance
T1059
Command and Scripting Interpreter
execution
T1036.005
Match Legitimate Resource Name or Location
defense-evasion
Guidance Disclaimer
The analysis, framework mappings, and incident response recommendations in this intelligence
item are derived from established industry standards including NIST SP 800-61, NIST SP 800-53,
CIS Controls v8, MITRE ATT&CK, and other recognized frameworks. This content is provided
as supplemental intelligence guidance only and does not constitute professional incident response
services. Organizations should adapt all recommendations to their specific environment, risk
tolerance, and regulatory requirements. This material is not a substitute for your organization's
official incident response plan, legal counsel, or qualified security practitioners.