Cisco Expressway: Steps to Patch and Hunt for Intrusions

Cisco Expressway: Steps to Patch and Hunt for Intrusions

Attackers exploited CVE-2026-76466, with a score of 9.8, to compromise Cisco Expressway devices belonging to a Russian organization and reach nodes on its internal network. F6 observed two internet-accessible devices that had been compromised and, from one of them, access to two other cluster nodes; the attackers gained root privileges and prepared scripts to explore the network.

The flaw allows unauthenticated execution of operating system commands as root if the attacker has network access to the device. According to SecurityWeek, it affects virtual and appliance versions of Expressway Core and Expressway Edge when they are clustered. The observed campaign targeted X15.5.3 devices.

1. Identify your Expressway devices and their versions

  1. Inventory Core and Edge devices, in both virtual and appliance formats. Record the installed version and whether each device is part of a cluster; prioritize nodes accessible from the internet. Expressway allows remote employees, partners, and external users to connect to corporate collaboration services.

  2. Compare each version with the fix. Cisco published its advisory on October 21, 2026, and says Expressway X15.5.20 fixes the vulnerability. For X14 and earlier versions, Cisco advises migrating to a fixed version. The following table summarizes the published guidance:

    Installed versionRecommended action
    X14 and earlierMigrate to a fixed version.
    X15.5.3 and earlierUpgrade to X15.5.20.

    Check the upgrade path that applies to your installation and use a fixed, compatible version. Do not assume a device is protected just because it is not running X15.5.3: Cisco's advisory covers the branches listed in the table.

2. Reduce exposure while arranging the patching

  1. Schedule the upgrade urgently and apply it to all affected Expressway devices, not just nodes that receive external connections. Cisco states that there is no workaround that fixes the vulnerability; network filtering recommended by F6 is a temporary measure, not a substitute for the patch.

  2. Restrict TCP/4371–4372 if you cannot patch yet. F6 recommends blocking connections to these ports at the firewall, allowing them only from the IP addresses of trusted cluster nodes. Apply the restriction while preparing the upgrade and verify that it does not leave access open from other networks.

3. Look for signs of compromise

  1. Check erlang.log on the Expressway devices for the string nodeup,'[email protected]'. F6 identifies it as an indicator of exploitation related to CVE-2026-76466. If it appears, also examine other logs and artifacts to confirm the intrusion: the string is an indicator to investigate, not standalone proof of the full scope.

  2. Search for the Python scripts identified by F6 under the path /mnt/harddisk/i{0-9}{4}[a-zA-Z]?/*.py. F6 says the attackers used custom scripts to scan the network, locate SMB resources, interact with SQL Server and WinRM, and examine GitLab repositories for credentials and other sensitive information.

  3. Compare activity against the network indicators: F6 published the addresses 66.179.31[.]32 and 18.130.47[.]116. Check relevant logs to determine whether there were associated connections. The absence of these indicators does not by itself rule out an intrusion: F6 also recommends reviewing other logs and artifacts.

“This incident demonstrates that compromise of a perimeter device can open a path for an attacker to an organization's critical systems.”

— F6, cybersecurity company

4. If you find signs of compromise, contain it and expand the investigation

  1. Isolate compromised systems and determine which devices and segments were reached. F6 observed a jump from an exposed Expressway device to two internal cluster nodes; therefore, limiting the review to perimeter devices would be insufficient.

  2. Investigate possible lateral movement across the network and in the services the scripts targeted: SMB resources, SQL Server, WinRM, and GitLab repositories. F6 reported attempts to extract credentials and other sensitive data. Review access and activity in those environments to establish what may have been exposed, and block any paths of further movement that you identify.

  3. Upgrade even after containment. Eliminating exposure does not establish whether exploitation occurred earlier. Cisco fixed the flaw with updated software; F6 recommends upgrading, auditing the environment, isolating affected systems, and determining the scope of the attack.

Mistakes that can leave the incident unresolved

Do not treat filtering as a patch: restricting TCP/4371–4372 only reduces connections while you are unable to upgrade, and Cisco offers no workaround that fixes the flaw. Also, do not review only the device exposed to the internet; the case investigated by F6 reached other cluster nodes on the internal network.

Do not close the alert based on a single indicator. The string in erlang.log warrants investigation, but F6 advises checking other logs and artifacts. Similarly, failing to find the published IP addresses or scripts at the specified path does not prove that no access occurred.

F6 did not attribute the operation to a known actor. The investigation should focus on evidence from your environment, the scope of the activity, and fixing all affected Expressway devices.

Sources: The Hacker News, iSpazio, F6, Cisco, SecurityWeek