Skip to content
All posts

F5 BIG-IP APM CVE-2026-123637: Critical Flaw, What to Do

November 22, 2026 · Team PlayCISO
TL;DR · the short version

F5 published CVE-2026-123637 on 18 November 2026: a stack-based buffer overflow in BIG-IP Access Policy Manager (APM), CVSS 3.1 9.8 and CVSS 4.0 9.3. F5 says undisclosed requests from a remote, unauthenticated attacker can crash the Traffic Management Microkernel (TMM), interrupting traffic while it restarts, and may lead to remote code execution. It applies only when APM is provisioned and a per-request policy containing both the Client Information agent and Endpoint Security agents is attached to a virtual server; it is a data plane issue, so restricting the management interface does not help. Affected lines are 21.1.0, 17.5.x and 17.1.x; F5 lists engineering hotfixes as the first unaffected builds and no workaround. CISA's SSVC entry says no exploitation; it is not in CISA KEV. F5 credits Sina Kheirkhah of watchTowr.

Put it into practice
Rank CVE-2026-123637 against the rest of your backlog
Score it by exposure, exploitation and impact: an internet-facing APM virtual server with the vulnerable policy goes first.
Open the prioritizer

F5 says a remote, unauthenticated attacker can crash BIG-IP's traffic engine, and possibly run code, through CVE-2026-123637 in BIG-IP APM. F5 published the flaw on 18 November 2026 in its CVE record and advisory K000163937: a stack-based buffer overflow scored CVSS 3.1 9.8 (CVSS 4.0 9.3). It only applies when APM is provisioned and a per-request policy with both the Client Information agent and Endpoint Security agents is attached to a virtual server. F5 lists engineering hotfixes and no workaround. No exploitation has been reported.

Who this is for: CISOs, network and remote-access teams who run F5 BIG-IP APM for VPN, SSO or endpoint posture checks, and anyone who owns the patch queue for edge devices.

Based on F5's CVE record and the CISA KEV catalog, read on 22 November 2026. F5's K-article did not load for us, so F5 facts here come from its CVE record. GBHackers' 21 November report was our lead. PlayCISO is not affiliated with F5 or watchTowr. We publish no exploit details; F5 has not disclosed the triggering requests.

Confirmed vs not yet known

F5 confirms (18 Nov): CVE-2026-123637, CWE-121, in BIG-IP APM 21.1.0, 17.5.x and 17.1.x. Undisclosed requests "can cause the Traffic Management Microkernel (TMM) to terminate and may lead to remote code execution (RCE)". Remote and unauthenticated. "There is no control plane exposure; this is a data plane issue only." Workarounds: none.

CISA's view (18 Nov SSVC): Exploitation none, Automatable yes, Technical Impact total. Not in the KEV catalog as of version 2026.11.20.

Credit: Sina Kheirkhah (SinSinology) of watchTowr, per F5.

Not established: any exploitation in the wild, any public proof of concept, and whether code execution has been demonstrated. F5 says the crash is certain and RCE is possible ("may lead to"); a TMM restart on its own does not show code execution. F5 lists no fixed base release, only hotfixes, in its CVE record.

Am I affected?

You are exposed only if all three are true:

  1. The APM module is provisioned on a BIG-IP running 21.1.0, 17.5.x or 17.1.x (and not yet on a hotfix build below).
  2. A per-request policy on that system includes both the Client Information agent and Endpoint Security agents.
  3. That policy is attached to a virtual server that the attacker can reach.

Because F5 calls it a data plane issue, management-interface restrictions do not help. The virtual server facing users, often an internet-facing VPN or SSO portal, is where the exposure is. F5 says versions past End of Technical Support were not evaluated: treat them as unknown, not safe.

Affected versions and hotfixes

BIG-IP APM lineFirst unaffected build (F5 CVE record)
21.1.0BIGIP-21.1.0.4.0.19.3-ENG
17.5.0 to 17.5.xBIGIP-17.5.1.11.0.61.12-ENG or BIGIP-17.5.1.10.0.80.35-ENG
17.1.0 to 17.1.xBIGIP-17.1.3.7.0.21.4-ENG or BIGIP-17.1.3.6.0.14.20-ENG

These are engineering hotfixes built on specific base releases (21.1.0.4, 17.5.1.11 or 17.5.1.10, 17.1.3.7 or 17.1.3.6), so match the hotfix to your exact installed build. GBHackers reports that F5's advisory lists other BIG-IP modules, BIG-IQ, F5OS, BIG-IP Next for Kubernetes and NGINX as not affected; we could not open the K-article to confirm that list. F5's CVE record names only BIG-IP APM.

Useful next, if this was relevant to you

Showing 1 of 3: CISO Readiness Scorecard

Timeline

DateEventSource
10 Nov 2026CVE-2026-123637 reserved by F5CVE.org
18 Nov 2026F5 publishes the CVE and advisory K000163937 (CVSS 9.8); credits watchTowrCVE.org, F5
18 Nov 2026CISA SSVC: Exploitation none, Automatable yes, Technical Impact totalCVE.org (CISA ADP)
20 Nov 2026Not in CISA KEV (catalog version 2026.11.20)CISA KEV
21 Nov 2026GBHackers reports the flawGBHackers

What security leaders should do this week

  • Find every BIG-IP with APM provisioned on 21.1.0, 17.5.x or 17.1.x, including disaster-recovery pairs, Virtual Edition and cloud instances. Note anything past End of Technical Support separately.
  • Check the per-request policies. List those that contain both the Client Information agent and Endpoint Security agents, and which virtual servers they are attached to. Those systems are the exposed ones.
  • Install the matching hotfix, internet-facing virtual servers first. CISA marks it automatable with total impact, even with no exploitation yet. Test APM policy behaviour after an engineering hotfix, then plan the move to a base release that includes the fix when F5 ships one.
  • Watch for TMM restarts. An unexplained TMM core or restart on an APM system is worth investigating now, not just logging. Keep the core files for F5 Support.
  • Record the decision. Score it with the SSVC calculator, check exploitation odds in EPSS lookup, and log any APM system you cannot patch this week in your risk register with an owner and a date. If it ever lands in KEV, the BOD 26-04 calculator gives your deadline.
  • Rehearse the bad day. The Edge VPN Intrusion Lab walks through the decisions when the remote-access box at your perimeter is the problem.

F5 also fixed two NGINX HTTP/2 flaws on the same day; see NGINX HTTP/2 RCE CVE-2026-123007. For the vendor view, see F5 in the Breach Index. CVE-2026-123637 is a vulnerability in BIG-IP, not a breach of F5.

Frequently asked questions

What is CVE-2026-123637?

CVE-2026-123637 is a critical stack-based buffer overflow (CWE-121) in F5 BIG-IP Access Policy Manager (APM), published by F5 on 18 November 2026. F5 says undisclosed requests can make the Traffic Management Microkernel (TMM) terminate and may lead to remote code execution, by a remote, unauthenticated attacker. F5 scores it CVSS 3.1 9.8 and CVSS 4.0 9.3.

Which BIG-IP configuration is vulnerable to CVE-2026-123637?

Per F5, only when BIG-IP APM is provisioned and a per-request policy that includes both the Client Information agent and Endpoint Security agents is applied to a virtual server. F5 says there is no control plane exposure: it is a data plane issue, so locking down the management interface does not remove the risk on an affected virtual server.

Which BIG-IP versions are affected and fixed?

F5's CVE record lists BIG-IP APM 21.1.0, 17.5.0 to 17.5.x and 17.1.0 to 17.1.x as affected. The first unaffected builds it lists are engineering hotfixes: BIGIP-21.1.0.4.0.19.3-ENG; BIGIP-17.5.1.11.0.61.12-ENG or BIGIP-17.5.1.10.0.80.35-ENG; and BIGIP-17.1.3.7.0.21.4-ENG or BIGIP-17.1.3.6.0.14.20-ENG. Pick the hotfix that matches your exact base build. Versions past End of Technical Support were not evaluated.

Is there a workaround for CVE-2026-123637?

F5 lists none ("Workarounds: None" in its CVE record). Until you can install the hotfix, the exposure depends on whether any virtual server carries a per-request policy with both the Client Information and Endpoint Security agents.

Is CVE-2026-123637 being exploited?

No exploitation has been reported in the sources we read. CISA's SSVC entry on the CVE record (18 November 2026) says Exploitation: none, Automatable: yes, Technical Impact: total. It was not in the CISA KEV catalog as of version 2026.11.20. F5 has not published the requests that trigger it.

Who found CVE-2026-123637?

F5 credits Sina Kheirkhah (SinSinology) of watchTowr.

Was F5 hacked again?

No. CVE-2026-123637 is a vulnerability in BIG-IP software, not a breach of F5. F5's confirmed breach is the 2025 nation-state theft of BIG-IP source code, tracked separately in the Breach Index.

Sources

Rank CVE-2026-123637 against the rest of your backlog
Score it by exposure, exploitation and impact: an internet-facing APM virtual server with the vulnerable policy goes first.
Open the prioritizer
Take this with you
Am I affected? · Affected versions and hotfixes · Timeline · What security leaders should do this week · Sources

Ready to practise the decisions these articles describe?

Run your first Cyber War Room free →

Pro