Threat intel5 min read
Attackers exploit critical Cisco Expressway flaw to gain root on clustered nodes
CVE-2026-76466 lets an unauthenticated attacker run commands as root on clustered Cisco Expressway. It is now exploited and listed by CISA. Upgrade to X15.5.20.
4Tify Team

Attackers are exploiting CVE-2026-76466, a critical flaw in clustered Cisco Expressway deployments that hands an unauthenticated attacker root on the device. CISA added it to its Known Exploited Vulnerabilities catalog on 20 November 2026 and gave US federal agencies until 23 November to act. Cisco shipped the fix a month ago, so any cluster still on X15.5.3 or older should be treated as urgent.
What happened
Cisco disclosed the bug on 21 October 2026 and rated it 9.8 out of 10. At the time it said it knew of no malicious use, and its advisory still says so.
That changed after exploit code appeared on GitHub in early November. Soon after, attackers used it against an organization in Russia. They first took over two internet-facing Expressway nodes running X15.5.3, then used one of them to reach two more cluster nodes on the internal network. Four devices were compromised in total.
With root on those boxes, the intruders ran custom Python scripts to map the network, look for open SMB shares, talk to SQL Server and WinRM, and search GitLab repositories. The aim was credentials and other sensitive data. Comments and fix notes in the scripts suggest they were written with help from an AI tool. Nobody has tied the activity to a known group.
Who is affected
The flaw sits in Expressway Series, both Expressway-C (Core) and Expressway-E (Edge), virtual and appliance. It only matters if the device is part of a cluster. Standalone nodes are not in Cisco's affected list.
| Release | Fix |
|---|---|
| X15.5.3 and earlier | Upgrade to X15.5.20 |
| X14 and earlier | Migrate to a fixed release |
Cisco lists no workaround.
How it works
Expressway nodes in a cluster share a database service and talk to each other over it. Cisco says the service relies on persistent cookies, and that when TLS Verify Mode is enabled for clustering, client certificates are not actually checked. Anyone holding those cookies who can reach the service can call remote procedure commands and run operating system commands as root. Cisco classifies the root cause as hard-coded credentials (CWE-798). That kind of weakness is the same on every vulnerable install, which makes a public exploit easy to reuse.
The cluster database uses TCP ports 4371 and 4372. If those ports are reachable from the internet or from user networks, the box is exposed.
What to do now
- Patch. Move every node in each cluster to X15.5.20 or later, following Cisco's cluster upgrade order. Check the running version under System > Status. The details are in Cisco's advisory.
- Fence the cluster ports. Until the upgrade is done, block TCP 4371 and 4372 at the firewall for everything except the IP addresses of the other cluster nodes. This cuts exposure but does not fix the flaw.
- Look for signs of compromise. Search the Erlang system log (erlang.log) for a
nodeupentry namingwombat@127.0.0.1, which marks a successful connection by the exploit. Check for Python scripts staged under/mnt/harddisk/. Compare the cluster's peer list against the nodes you expect. - If you find a hit, assume the network is in play. In the confirmed case the attackers moved from Expressway to file shares, databases and source code. Rotate credentials that were reachable from the cluster and review SMB, SQL Server, WinRM and GitLab logs for activity coming from Expressway addresses.
Edge appliances like Expressway are easy to forget once they work. Our Attack Surface Management service keeps an inventory of what you expose to the internet, including which software version each host runs, so a node left on an exploited release shows up before someone else finds it.