• T1190 – Exploit Public-Facing Application
  Attackers exploit the public-facing FortiManager application via a missing authentication flaw. This vulnerability allows unauthorized attackers to execute arbitrary code on FortiManager by sending specially crafted requests, gaining initial access to the system and enabling control over FortiGate devices connected to the network.

• T1078 – Valid Accounts
  The threat actors leverage valid certificates on unauthorized FortiManager and FortiGate devices, allowing them to register these devices on exposed FortiManager instances. By mimicking legitimate access, the attackers avoid raising immediate security alerts and maintain a low profile for further exploitation and lateral movement within the network.

• T1036 – Masquerading
  Attackers register rogue FortiManager devices under misleading names (e.g., “localhost”) and legitimate-seeming serial numbers (e.g., FMG-VMTM23017412). This technique helps obscure threat actor activity within FortiManager logs and console, allowing the attacker’s device to appear as if it is part of the legitimate infrastructure.

• T1041 – Exfiltration Over C2 Channel
  Exfiltration of FortiManager and FortiGate configuration files occurs over encrypted Command and Control (C2) channels, leveraging HTTPS to avoid detection by security tools. The threat actor UNC5820 has been observed using specific IP addresses to exfiltrate compressed files containing sensitive configuration information, user credentials, and device data.

• T1587.003 – Develop Capabilities: Digital Certificates
  Attackers leverage valid digital certificates on FortiManager and FortiGate devices to masquerade malicious activities as legitimate. With these certificates, unauthorized devices can connect to FortiManager, bypassing certain security configurations and enabling persistent access to compromised networks.

• T1562.001 – Impair Defenses: Disable or Modify Tools
  Attackers modify FortiManager configuration to evade detection. By using commands such as fgfm-deny-unknown, attackers can prevent detection of unauthorized devices. This adjustment allows attackers to sustain their unauthorized access, mitigating the chances of detection during ongoing operations.

• T1027 – Obfuscated Files or Information
  Attackers use gzip compression on the /tmp/.tm archive, which stores exfiltrated configuration data, to obfuscate and minimize visibility of extracted data. This technique reduces the file’s detection footprint, making it harder to identify during data exfiltration stages.

• T1040 – Network Sniffing
  While not directly observed in this incident, the configuration data exfiltrated includes sensitive details like IPs and credentials. This could indicate an intention to use network sniffing techniques or other credential-monitoring tactics to further penetrate or maintain persistence in the target network.

V. Immediate Recommendations

• Install Security Updates:
  • Fortinet has solved CVE-2024-47575 with fixes. To address the found security flaw and reduce the risk of active exploitation, organizations should give top priority to installing these updates on all FortiManager instances, including on-premises and cloud-based.

• Monitor for Compromise Indicators (IoCs):
  • Check network traffic and system logs often for known IoCs linked to this attack, such as file paths, flagged IP addresses, MD5 hash values, and log entries that might point to exploitation (see to the IoCs section for references). To improve detection capabilities, incorporate these IoCs into your SIEM or IDS/IPS.

• Establish an Incident Response Plan:
  • Create or revise an incident response plan that includes steps for handling FortiManager vulnerability exploitation. Make sure your reaction team is equipped and trained to deal with any possible Fortinet system breaches.

• Isolate Compromised Systems:
  • Isolate compromised systems right away to stop additional access or harm if any indications of compromise are found. Notify the affected parties and carry out a comprehensive investigation, eliminating any malware or backdoors.

VI. IOCs (Indicators of Compromise)

• T1190 – Exploit Public-Facing Application
  • The attackers exploit a vulnerability in the Zimbra Collaboration Suite, a public-facing application, by sending specially crafted emails that trigger command execution on the server.
• T1505.003 – Server Software Component: Web Shell
  • The attackers create a web shell on the compromised server by concatenating base64-encoded commands from the CC field of the emails, allowing persistent remote access.
• T1059.004 – Command and Scripting Interpreter: Unix Shell
  • The attackers execute shell commands on the server by exploiting the input validation flaw, enabling them to control the system via the web shell.
• T1071.001 – Application Layer Protocol: Web Protocols
  • The attackers use HTTP requests with specially crafted cookies (JSESSIONID and JACTION) to communicate with the web shell, establishing a command-and-control channel.
• T1105 – Ingress Tool Transfer
  • Through the web shell, the attackers download and execute additional malicious code or files onto the compromised server.
• T1132.001 – Data Encoding: Standard Encoding
  • The attackers use base64 encoding to encode malicious commands and payloads within the email CC fields and cookies to obfuscate the data and evade detection.
• T1036.005 – Masquerading: Match Legitimate Name or Location
  • The attackers send spoofed emails that appear to come from Gmail, leveraging trusted sources to bypass initial security checks.

V. Recommendations

• Patch Management
  • Ensure that all Zimbra email server installations, including Zimbra 9.0.0 Patch-41, Zimbra 10.0.9, and Zimbra 10.1.1 (Daffodil), are updated with the latest patches addressing CVE-2024-45519. Systems still running Zimbra 8.8.15, which has received a one-time patch past its EOL, should be prioritized for patching. Regularly monitor for new security updates and apply them as soon as they are released.
• Monitoring and Logging
  • Implement comprehensive monitoring and logging to detect suspicious activities targeting the Zimbra postjournal service. Focus on identifying unusual email patterns, base64-encoded commands, or abnormal execution of commands through the postjournal service. Regular log reviews can help catch early signs of exploitation.

• Access Control
  • Properly configure Zimbra’s “mynetworks” parameter to restrict access to trusted IP ranges only. If the postjournal service is not required for your organization’s operations, consider disabling it to reduce the attack surface, especially in environments where patching may be delayed
• Service Management
  • Ensure that optional services like postjournal, which is not enabled by default, remain disabled unless explicitly needed. On systems where postjournal is unnecessary, consider removing or disabling it entirely to minimize potential vulnerabilities.
• Vendor Communication
  

  • Establish regular communication with Zimbra to stay informed about the latest security advisories, patches, and best practices. Regularly check the Zimbra Security Center and set up notifications to receive updates on new vulnerabilities and security patches promptly.

VI. IOCs (Indicators of Compromise)