CVE-2026-76466: Cisco Expressway Root RCE Exploited Weeks After Patch

A critical root RCE in clustered Cisco Expressway devices, patched on October 21, has now been exploited in the wild according to F6. Every automated signal on the CVE still shows its day-one state. Here is who is affected, how to upgrade and what to look for in your logs.

CVE-2026-76466: Cisco Expressway Root RCE Exploited Weeks After Patch

Cisco patched CVE-2026-76466 in Cisco Expressway on October 21, 2026, and rated it Critical with a CVSS 3.1 score of 9.8. The flaw lets an unauthenticated remote attacker run operating system commands as root on clustered Expressway-C and Expressway-E devices. Cisco lists no workaround. The fix is an upgrade to X15.5.20, and X14 and earlier must migrate to a fixed release.

At release, Cisco PSIRT said it was not aware of public announcements or malicious use. That changed in November. Russian security company F6 told CNews on November 18 that it had investigated the first observed exploitation of the flaw, against a company in Russia, shortly after a public proof-of-concept appeared. THN reported the same findings on November 20. The exploitation claim rests on F6's account, relayed by these two outlets; F6's own report was not linked in either article.

According to F6, attackers first took over two internet-facing Expressway devices running X15.5.3, then used one of them to reach two more cluster nodes on the internal network. From there they prepared custom Python tools to scan the network, find SMB shares, talk to SQL Server and WinRM, and search GitLab repositories for credentials. F6 also said comments in the scripts suggest an AI assistant helped write them. The attack has not been attributed to a known group.

How it works

The bug sits in the database service feature of Expressway. Cisco gives two causes. The first is the use of persistent cookies. The second is that when TLS Verify Mode is set to Enabled for clustering, client certificates are not actually validated.

Put together, a party that holds those cookies can connect to the database service and call specific remote procedure call (RPC) commands, which run as root on the underlying system. Cisco classes the weakness as CWE-798, Use of Hard-coded Credentials, and the advisory ID itself carries the label "static-cred".

The condition that matters most is clustering. Cisco says the flaw affects Expressway-C and Expressway-E, virtual and appliance, only when they run a vulnerable release and are configured in a cluster. F6 told CNews that the log entry it found matches the known exploitation path, and recommended limiting TCP ports 4371 and 4372 to trusted cluster peers when an upgrade cannot happen right away. That port guidance comes from F6, not from Cisco's advisory.

Who is affected

Any clustered Expressway-C or Expressway-E on X15.5.3 or earlier is exposed. Cisco's fixed-release table is short:

Cisco Expressway release

First fixed release

X14 and earlier

Migrate to a fixed release

X15.5.3 and earlier

X15.5.20

A few caveats:

  • Standalone devices. Cisco ties this CVE to clustered setups. The same X15.5.20 release also fixes eight other Expressway CVEs (CVE-2026-76490 to CVE-2026-76497) from Cisco's October hardening advisory, which apply regardless of configuration, so non-clustered devices still need the upgrade.
  • X14 has no patch. There is no X14 fix to apply; those devices move to a fixed release.
  • Version lists differ. The CVE record lists affected versions from X8.1 through X15.5.3 under the older product name "TelePresence Video Communication Server (VCS) Expressway". Read that list as the version range, and Cisco's table as the action.
  • Already compromised? Patching does not undo a breach. F6's devices ran X15.5.3, the last vulnerable X15 release, so recent versions are not a sign of safety.

The signals were frozen on day one

Every automated signal on this CVE still describes October 21, while the attack happened in November. The scores were right when they were set. Nothing in them has moved since.

Date (2026)

What happened

Oct 21

Cisco publishes the advisory and fixes; PSIRT knows of no malicious use. CISA's SSVC entry in the CVE record is stamped the same day: Exploitation none, Automatable yes, Technical impact total.

Oct 26

Two public repositories appear offering offline detectors for the flaw.

Nov 5

A public repository appears describing a Python exploit and fleet detector.

Nov 18

F6 tells CNews it investigated the first exploitation, which followed a public proof-of-concept.

Nov 20

THN reports F6's findings.

As of today, the CVE record still carries the October SSVC value of no known exploitation and no KEV data. EPSS sits at 0.4%, a figure built to predict exploitation, not to report it. The vendor advisory is still version 1.0 with the original exploitation statement.

The practical point is that two of the three SSVC inputs, Automatable and Total impact, were already at their worst on day one. The only input that stood between this CVE and the top of a patch queue was "no known exploit", and that is the input most likely to go stale. When a flaw is network-reachable, unauthenticated and root-level, a quiet exploitation field is a reason to check the news, not a reason to wait.

Reading it on VulnTracker

The CVE page shows the October picture accurately, and the Exploits tab is where the November story starts to show.

Field

Value

Note

CVSS

9.8 Critical

CVSS 3.1 from Cisco as CNA (AV:N/AC:L/PR:N/UI:N)

EPSS

0.4%

A probability estimate, not an exploitation report

SSVC

No known exploit · Automatable · Total impact

Matches CISA's October 21 entry

Status

Awaiting Analysis

NVD has not added its own assessment yet

CWE

CWE-798

Added in a maintenance update after publication

Published / updated

Oct 21, 2026 / Oct 22, 2026

The Timeline tab shows the publish date as 10/22

Exploits tab

3 public repositories

1 described as an exploit, 2 as offline detectors

Vendor advisories

Cisco (Critical, vendor fix) and CERT-Bund WID-SEC-W-2026-4105 (High)

Data notes:

  • Exploits tab vs SSVC. The Exploits tab lists three repositories while SSVC says "No known exploit". These are not in conflict: SSVC tracks exploitation in the wild, and the Exploits tab lists public code. Two of the three are detectors, not exploits. We list them here and do not link them.
  • Remediation panel. The Remediation block says no fixed version or vendor advisory is recorded yet, while the Vendor advisories block shows Cisco's advisory with a vendor fix. Use Cisco's table above as the fix reference.
  • Severity labels. CERT-Bund rates its advisory High; Cisco rates the CVE Critical. CERT-Bund's entry covers several Expressway flaws together.
  • Exploitation. In-the-wild exploitation reported by F6 is not yet reflected in the SSVC field. Check the Newsroom tab for the reports.

What to do

  1. Inventory. List every Expressway-C and Expressway-E, its release, and whether it is in a cluster.
  2. Upgrade. Move X15 devices to X15.5.20. Plan the migration for anything on X14 or earlier, since there is no X14 fix.
  3. Restrict cluster traffic until then. F6 recommends allowing TCP 4371 and 4372 only from trusted cluster peer addresses. Cisco itself lists no workaround, so treat this as a stopgap.
  4. Hunt before you close the ticket. Check erlang.log on each device for a nodeup entry naming [email protected], which F6 ties to the exploitation path. F6 and THN also list two source IPs, 66.179.31[.]32 and 18.130.47[.]116, and Python scripts staged under /mnt/harddisk/.
  5. Assume reach beyond the box. If you find signs of compromise, look at what the device could talk to. In F6's case that meant other cluster nodes, SMB shares, SQL Server, WinRM and GitLab.
  6. Rotate what was reachable. Credentials stored on or reachable from a compromised Expressway, including service accounts and repository tokens, should be treated as exposed.

Takeaway

CVE-2026-76466 went from "no known malicious use" to a reported four-device breach in under a month, and the public exploit came first. A clustered perimeter device with an unauthenticated root flaw belongs at the top of the queue on the day the advisory lands, whatever the exploitation field says. If you run Expressway, the work is the upgrade plus a look at the logs.

If you are not sure whether Expressway sits somewhere in your estate, check CVE-2026-76466 against your stack on VulnTracker and follow its status as the exploitation data catches up.

References