<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/"><channel><title>DSC - CraftedSignal Threat Feed</title><link>https://feed.craftedsignal.io/products/dsc/</link><description>Trending threats, MITRE ATT&amp;CK coverage, and detection metadata. Fed continuously.</description><generator>Hugo</generator><language>en</language><managingEditor>hello@craftedsignal.io</managingEditor><webMaster>hello@craftedsignal.io</webMaster><lastBuildDate>Fri, 19 Jun 2026 12:48:20 +0000</lastBuildDate><atom:link href="https://feed.craftedsignal.io/products/dsc/feed.xml" rel="self" type="application/rss+xml"/><item><title>Azure VM Extension CRUD from Unusual Source ASN</title><link>https://feed.craftedsignal.io/briefs/2026-06-azure-vm-extension-crud-unusual-source/</link><pubDate>Fri, 19 Jun 2026 12:48:20 +0000</pubDate><author>hello@craftedsignal.io</author><guid isPermaLink="true">https://feed.craftedsignal.io/briefs/2026-06-azure-vm-extension-crud-unusual-source/</guid><description>Threat actors are performing create, read, update, or delete (CRUD) operations against Azure VM or VM Scale Set extensions (e.g., CustomScript, DSC) from an anomalous source Autonomous System (AS) number, enabling high-privilege code execution and persistence on guest operating systems (SYSTEM on Windows, root on Linux) by abusing compromised Azure identities.</description><content:encoded><![CDATA[<p>This brief details a threat identified by Elastic, focusing on the abuse of Azure Virtual Machine (VM) and VM Scale Set (VMSS) extensions. Threat actors can perform create, read, update, or delete (CRUD) operations on these extensions, such as <code>CustomScript</code> or <code>Desired State Configuration (DSC)</code>, from an unusual source Autonomous System (AS) number. These extensions execute with high privileges (SYSTEM on Windows, root on Linux) on the guest operating system, making them a prime target for initial code execution, maintaining persistence, or defense evasion. This technique allows adversaries to run arbitrary commands, install malware, or modify system configurations without direct login, leveraging compromised Azure credentials or identities. The detection specifically targets activity originating from networks not historically associated with managing a given extension resource, while excluding benign first-party Microsoft automation.</p>
<h2 id="attack-chain">Attack Chain</h2>
<ol>
<li><strong>Initial Access</strong>: Attacker obtains valid Azure credentials (e.g., user account, service principal) through methods such as phishing, credential stuffing, or exploiting a misconfiguration.</li>
<li><strong>Privilege Escalation/Lateral Movement (Azure Plane)</strong>: Attacker identifies a target Azure subscription or resource group with permissions to manage VM or VM scale set extensions.</li>
<li><strong>VM Extension CRUD Operation</strong>: Attacker uses the compromised credentials to perform a <code>WRITE</code> (create/update), <code>DELETE</code>, or <code>READ</code> operation against an Azure VM or VMSS extension. This operation originates from an AS number not typically observed for managing that specific resource.</li>
<li><strong>Code Execution (Guest OS)</strong>: If a <code>WRITE</code> operation is performed using extensions like <code>CustomScript</code> or <code>DSC</code>, the malicious script or command embedded in the extension definition is executed on the target VM's guest OS with SYSTEM (Windows) or root (Linux) privileges.</li>
<li><strong>Persistence/Defense Evasion</strong>: The executed code establishes persistence mechanisms, such as new services, scheduled tasks, or modifying existing configurations, or removes security agents to evade detection.</li>
<li><strong>Internal Reconnaissance &amp; Data Exfiltration</strong>: With high privileges on the VM, the attacker performs internal network reconnaissance, collects sensitive data, and prepares for exfiltration to attacker-controlled infrastructure.</li>
<li><strong>Impact &amp; Follow-on Activity</strong>: The attacker might deploy ransomware, conduct further lateral movement across the internal network, or maintain long-term access for data theft.</li>
</ol>
<h2 id="impact">Impact</h2>
<p>Successful exploitation of Azure VM extensions grants attackers SYSTEM or root-level privileges on target virtual machines, leading to severe consequences. This can result in unauthorized code execution, installation of persistent backdoors, and the ability to disable security controls. Organizations can face significant data breaches, potential ransomware deployment, and complete compromise of critical cloud infrastructure. The impact extends to business disruption, regulatory non-compliance, and substantial financial and reputational damage. While specific victim counts are not available for this general technique, highly privileged access on cloud assets is consistently associated with the most severe incident types.</p>
<h2 id="recommendation">Recommendation</h2>
<ul>
<li>Deploy the provided Sigma rules to your SIEM, focusing on Azure Activity Logs (<code>category: cloud</code>, <code>product: azure</code>).</li>
<li>Enable comprehensive logging for Azure Activity Logs across all subscriptions to capture <code>MICROSOFT.COMPUTE/VIRTUALMACHINES/EXTENSIONS</code> events.</li>
<li>Implement a baseline of expected <code>source.as.number</code> values for all Azure VM/VMSS extension management activities and create an allowlist for known, legitimate ASNs (e.g., CI/CD pipelines, internal management networks).</li>
<li>Review <code>azure.activitylogs.identity.authorization.evidence.principal_id</code> and <code>...principal_type</code> fields in alerts to determine the legitimacy and permissions of the principal performing the operation.</li>
<li>Integrate endpoint detection and response (EDR) telemetry (e.g., <code>process_creation</code> events from <code>WaAppAgent.exe</code> or <code>walinuxagent</code>) on Azure VMs to correlate with <code>WRITE</code> extension operations for script execution.</li>
</ul>
]]></content:encoded><category domain="severity">high</category><category domain="type">advisory</category><category>cloud</category><category>endpoint</category><category>azure</category><category>azure-activity-logs</category><category>threat-detection</category><category>execution</category><category>persistence</category></item></channel></rss>