F5 rates six of 12 BIG-IP advisories critical and nine hotfix only; BleepingComputer lists five critical
F5's twelve BIG-IP and BIG-IQ articles of 18 November 2026 score six Critical on CVSS 3.1 and print only engineering hotfix builds as the fix for nine. BleepingComputer lists five critical flaws and no page this site read says any is exploited.
By Parminder Kumar Sharma · LinkedIn · · 23 min read
Six of twelve are critical, nine are hotfix only, and BleepingComputer lists five
F5's index article for its out-of-band security notification of 18 November 2026 lists 22 articles: ten for NGINX and twelve for BIG-IP, BIG-IQ and related products (counted from the saved index; derived). This site read all twelve of the second group on 20 November 2026, between 15:23 and 15:25 UTC. Counted from each article's own table (derived): six are Critical on CVSS 3.1, all at 9.8, and six are High, from 7.5 to 8.8. On CVSS 4.0, seven are Critical (six at 9.3, one at 9.2) and five are High, all at 8.7. None is below High. Nine of the twelve print only engineering hotfix builds as the fix, and three print release versions. All six that are Critical on CVSS 3.1 are among the nine.
BleepingComputer's report of 20 November calls them "critical BIG-IP RCE flaws" in a headline that also says "emergency patches", and says "All five security issues are rated critical severity". It names five CVE ids as critical, and one more, a SAML flaw, as high severity. F5's pages count six Critical on CVSS 3.1. The one missing from BleepingComputer's five, CVE-2026-120722 in K000163945, a flaw tied to the DNS::scrape iRule command, is not among BleepingComputer's ids (derived by comparing the two). BleepingComputer links six of the twelve F5 articles and names none of the other six's CVE ids.
What that does not establish. The word exploit appears on none of the twelve pages, the index or the hotfix note (derived by searching the saved text), so the F5 pages are silent on exploitation: they do not say it is happening and they do not say it is not. BleepingComputer writes that the company "hasn't flagged any of these issues as exploited in attacks". No page says how many devices run the affected component, policy or iRule command. No page prints a date for the next regular BIG-IP release. And a count of critical articles is not a count of exposed systems: each article names a condition, and the conditions differ.
Whose words these are. F5 sells BIG-IP, so its advisories are its own account of its own products. BleepingComputer is a news site and a secondary source here. Shadowserver, whose counts BleepingComputer relays, is a nonprofit that counts devices. This site did not read Shadowserver's dashboard. It names no individual: the F5 pages credit people, and only organisations appear here. Every figure this site worked out is labelled derived.
Twelve articles: five APM, five iRule, two iControl
Grouped by component (derived from the tables): five articles concern BIG-IP APM (K000163937, K000163938, K000163942, K000163943, K000163944), five concern an iRule command (K000163939, K000163940, K000163941, K000163945, K000163946), and two concern the iControl interfaces, REST (K000163929) and SOAP (K000163936). The diagram gives each article's component, its two scores and the fix type, in article order. Amber rows are the six Critical on CVSS 3.1.
The two scoring systems do not rank the twelve alike (derived from the printed scores). On five articles CVSS 4.0 is higher than 3.1: iControl REST by 1.1 (8.1 against 9.2, which moves it from High to Critical) and four APM articles by 1.2 (7.5 against 8.7). On the other seven 4.0 is lower: by 0.5 on each of the six that are Critical on both (9.8 against 9.3), and by 0.1 on iControl SOAP (8.8 against 8.7). So the word critical depends on the version of the score: six articles on 3.1, seven on 4.0. The iControl REST article is the only one whose band changes between them. Its Description is that REST tokens "may not expire as expected in all scenarios", and its Impact is that an attacker "who has obtained a token" may be able to extend its expiry lifetime unless the token owner deletes it. The page does not say how a token would be obtained.
F5 files seven of the twelve as CWE-121, a stack based buffer overflow (the five iRule articles, the APM per-request policy article and the APM form-based Logon Page article), three as CWE-191, an integer underflow (the three APM SAML and policy trace articles), one as CWE-1463, insufficient session timeout value (iControl REST), and one as CWE-93, improper neutralization of CRLF sequences (iControl SOAP). The classification does not sort by impact: the Logon Page article is a stack based buffer overflow like the iRule ones, but its printed impact is denial of service only and its scores are High 7.5 and 8.7.
What each article says has to be true first
Ten of the twelve describe a configuration: a module provisioned, a policy of a particular kind on a virtual server, an iRule command in use, or a database key switched on. The table copies the condition from each article's Description in short form, with exact quotations in marks, and the Impact as printed. F5's text says "undisclosed requests" for the trigger in the ten data plane articles; no page says what they are, and this site prints no trigger detail because the pages print none.
The condition and the impact each of the twelve F5 articles prints, as read on 20 November 2026. Wording shortened; quotations in marks are exact.
| Article and CVE | Condition in the Description | Impact as printed |
|---|---|---|
| K000163929 CVE-2026-120718 | Tokens "may not expire as expected in all scenarios" | An attacker who has obtained a token may be able to extend its expiry lifetime unless the token owner deletes it. No data plane or control plane statement. |
| K000163936 CVE-2026-125274 | "An authenticated iControl SOAP user with at least Guest access" | May allow an authenticated attacker with network access to execute arbitrary system commands, create or delete files, or disable services. Control plane only; no data plane exposure. |
| K000163937 CVE-2026-123637 | APM provisioned; a per-request policy with the Client Information agent and Endpoint Security agents applied to a virtual server | TMM terminates; denial of service; "may allow the attacker to perform RCE". Data plane only. |
| K000163938 CVE-2026-125272 | APM licensed and provisioned; a virtual server with the APM access profile and a per-request policy containing a form-based Logon Page | TMM terminates; denial of service. Data plane only. |
| K000163939 CVE-2026-120285 | An HTTP profile and an iRule containing the PSC::ip_address command on a virtual server | TMM terminates; denial of service; may allow RCE. Data plane only. |
| K000163940 CVE-2026-120292 | An HTTP profile and an iRule using the MR::message route command on a virtual server | TMM terminates; denial of service; may allow RCE. Data plane only. |
| K000163941 CVE-2026-116953 | An iRule that processes client traffic using the GENERICMESSAGE::route command on a virtual server | TMM terminates; denial of service; may allow RCE. Data plane only. |
| K000163942 CVE-2026-126822 | APM provisioned; a virtual server whose access profile is a SAML Service Provider (SP) | TMM terminates; denial of service. Data plane only. |
| K000163943 CVE-2026-126336 | APM provisioned; an APM access profile with a SAML SP agent with encrypted assertions enabled | TMM terminates; denial of service. Data plane only. |
| K000163944 CVE-2026-121812 | APM provisioned; an access profile and a per-request policy; the tmm.access.policytrace database key enabled, which is disabled by default | TMM terminates; denial of service, by a remote authenticated attacker. Data plane only. |
| K000163945 CVE-2026-120722 | An iRule that processes client traffic using the DNS::scrape command on a virtual server | TMM terminates; denial of service; the attacker "may be able to perform RCE". Data plane only. |
| K000163946 CVE-2026-121805 | An iRule that processes client traffic using the PEM::subscriber command on a virtual server | TMM terminates; denial of service; "may perform RCE". Data plane only. |
Three patterns stand out (derived). All five iRule articles are tied to one named iRule command, and the pages do not say how many deployments use any of them. The APM articles are tied to a policy type or agent, not to APM being provisioned alone: a per-request policy with two agents, a form-based Logon Page, a SAML Service Provider, encrypted assertions, or a policy trace key. One article, K000163944, needs a database key that F5 says is disabled by default, and that article's attacker is described as remote and authenticated, where nine of the other articles say remote and unauthenticated (derived by reading the Impact lines). F5 prints a command to check the key; it is in the technical section below.
The two iControl articles are a different kind. They concern the management interfaces rather than traffic handling, and F5's Impact line for SOAP says there is no data plane exposure and that it is a control plane issue only. That article carries the longest mitigation text of the twelve, and it is also the only one that lists BIG-IQ Centralized Management 8.4.0 to 8.4.2 as vulnerable, with a fix at 8.4.2.3.
"May lead to remote code execution" on six, denial of service on four
Reading the Description and Impact lines (derived counts): six articles say the flaw "may lead to remote code execution (RCE)", or may allow the attacker to perform it. They are the five iRule articles and the APM per-request policy article, and they are the same six that are Critical on CVSS 3.1. Four say denial of service and nothing more: the Logon Page, the SAML service provider, the encrypted assertions and the policy trace articles. One concerns token expiry and one concerns commands run by a Guest role user.
For the ten data plane articles F5 prints the same two sentences. "Traffic is disrupted while the TMM process restarts." and "There is no control plane exposure; this is a data plane issue only." The first is a statement about what happens to traffic: the process that handles it restarts. The second says the management side is not involved. A reader who takes "may lead to remote code execution" as a finding that code runs has read more than the page says; the verb is "may", and no article says code execution has been shown. A reader who reads only "denial of service" has also read less than the page says on the six that carry the RCE sentence. The articles do not describe how likely either outcome is.
The word restart matters to the one mitigation F5 prints for the encrypted assertions article, which is about availability: F5 recommends configuring BIG-IP systems with high availability "to lessen the impact of the vulnerability", and points to device service clustering and the heartbeat failsafe. That is a way to ride out a restart, not a fix, and the same article lists release versions as its fix.
BleepingComputer's words against the twelve pages
The diagram sets BleepingComputer's wording, as saved, beside what the F5 pages print on eight points. Where they differ, both are shown and none is resolved here.
The count. BleepingComputer says five of the flaws are critical and one is high severity. F5's pages carry six Critical on CVSS 3.1. The one missing from BleepingComputer's list is the DNS::scrape iRule article (K000163945, CVE-2026-120722). BleepingComputer's report is about the six articles it links; the other six, among them the two iControl articles, three of the APM ones and the DNS::scrape iRule one, are not named in it. It is a news report on part of the notification, and its headline says critical RCE flaws without saying which.
The wording. BleepingComputer says unauthenticated attackers "can exploit them to terminate" the traffic microkernel and "execute arbitrary code remotely on systems with vulnerable configurations". Its own quotation of F5 is hedged: "Traffic is disrupted while the TMM process restarts. This vulnerability allows a remote unauthenticated attacker to cause a denial-of-service (DoS) on the BIG-IP system and may perform RCE." The F5 articles use "may" on the code execution step for all six that mention it. The difference is between BleepingComputer's summary sentence and F5's page, in one report. This site notes it and does not rule on which wording a reader should carry.
Mitigation. BleepingComputer writes that "F5 said it has no mitigation measures for these flaws". That matches the six critical articles, whose Mitigation field reads None. It does not describe the whole notification: two more articles read None, and four print a mitigation (derived: eight of twelve say None). The four are REST token scripts, restricting SOAP access, removing the SAML agent and restricting the virtual server, and high availability.
The counts of devices. BleepingComputer writes that Shadowserver tracks "more than 177,000 BIG-IP instances" and "over 14,600 IP addresses with BIG-IP APM fingerprints", and adds that "there is no information on how many are honeypots or have already been patched". The ratio of the two figures is about 8.2 per cent (derived: 14,600 divided by 177,000 is 0.0825), but both are floors ("more than", "over"), they are fingerprints of devices, and the second counts BIG-IP APM, not the policy types the APM articles name. A fingerprint says a device answers like an APM; it does not say which policy sits on which virtual server. The same goes for iRule commands, which no external count can see. These are device counts, not counts of vulnerable configurations. This site did not read the dashboard, so the numbers are BleepingComputer's relay.
The history. BleepingComputer also writes that CISA has flagged seven F5 vulnerabilities as actively exploited since November 2021, that four were also used by ransomware gangs, and that F5 patched a critical BIG-IP APM flaw in September that is being exploited in zero day remote code execution attacks. These statements are about earlier flaws. This site did not read CISA's catalogue or any F5 page about the September flaw, and none of the twelve pages mentions it. They tell a reader how attackers have treated F5 products before. They do not bear on whether any of these twelve is exploited.
Hotfix or release version: what the Fixes introduced in column prints
Every one of the twelve tables lists a Fixes introduced in column. In nine articles it holds engineering hotfix (EHF) builds for BIG-IP, and in the five iRule articles it also holds hotfixes for BIG-IP CNF and BIG-IP Next for Kubernetes, from F5's Artifact Repository. In three articles, iControl REST, iControl SOAP and the APM encrypted assertions article, it holds release versions: 21.1.0.2 or 21.1.0.3, 17.5.1.10 and 17.1.3.6, and for BIG-IQ 8.4.2.3 on the SOAP article. Those three are also the three that carry a "Final" banner (derived), which says F5 will no longer update the article. A Final banner is a statement about the article, not about the reader's system: the same page still lists the versions known to be vulnerable.
The gap. The SSL Orchestrator row of the policy trace article, K000163944, lists 17.5.0 and 17.1.0 to 17.1.3 as vulnerable, and under fixes lists two 17.1.3 hotfix builds and none for 17.5 (derived from the saved row; the saved text does not keep cell borders, so a reader should check the live page). The page's own rule is that when the Fixes introduced in column lists no version for a branch, "no update candidate currently exists for that branch" and F5 recommends upgrading to a version with the fix. The same article's APM row lists 17.5.1.10 and 17.5.1.11 hotfix builds, so the APM and SSL Orchestrator rows differ.
Why hotfixes. The pages say that F5 "has fixed this issue in an engineering hotfix" and point to the MyF5 Downloads page and the hotfix note. They do not say why a hotfix is used instead of a release, and they do not give a date for the next regular release. They do say what a hotfix is not: every article with a hotfix says engineering hotfixes "do not undergo the extensive QA assessment of scheduled software releases" and that F5 offers them with "no warranty or guarantee of usability". That is F5's disclaimer, and it matters to anyone who runs change control on a load balancer in front of production.
Choosing the build. The hotfix note, K000164239, applies to TMOS v21.1.x, v17.5.x and v17.1.x and says the right build "depends on the BIG-IP version or EHF currently installed". It names five builds, four for the 17 lines and one for 21.1 (derived). The table below restates it in plain words. BIG-IQ is not covered, and the note says nothing about CNF or Next for Kubernetes.
Which hotfix build to install, as K000164239 prints it, read on 20 November 2026. The note lists one spelling with a hyphen after BIG, copied as printed.
| If you are running | Build to install | Branch |
|---|---|---|
| BIGIP-17.1.3.6.0.9.20-ENG | BIGIP-17.1.3.6.0.14.20-ENG | 17.1.x |
| The 17.1.3.7 base release or an earlier base release | BIGIP-17.1.3.7.0.21.4-ENG | 17.1.x |
| BIGIP-17.5.1.10.0.7.35-ENG or BIG-IP-17.5.1.10.0.37.35-ENG | BIGIP-17.5.1.10.0.80.35-ENG | 17.5.x |
| The 17.5.1.11 base release or an earlier base release | BIGIP-17.5.1.11.0.61.12-ENG | 17.5.x |
| The 21.1.0.4 base release or an earlier base release | BIGIP-21.1.0.4.0.19.3-ENG | 21.1.x |
| A custom engineering hotfix build | Contact F5 Support to request a new engineering hotfix | any |
A trap in the numbers (derived from F5's rule that a fix applies to later releases in a branch). The release version fixes for the three Final articles, 17.1.3.6 and 17.5.1.10, sit on the same bases as hotfix builds for the other nine. A system on 17.1.3.6 or 17.5.1.10 or later has, by the page's rule, the three release version fixes, and still needs a hotfix build for the nine. A system on 21.1.0.4, the base named in the hotfix note, is later than 21.1.0.2 and 21.1.0.3 and so has them too, and still needs the 21.1.0.4 hotfix. The pages do not say it in those words, so this is this site's reading of the tables.
The twelve tables as printed, and the one check F5 gives a command for
F5's tables repeat one set of branches. For BIG-IP the versions known to be vulnerable are 21.1.0 (branch 21.x), 17.5.0 to 17.5.1 and 17.1.0 to 17.1.3. The exceptions are SSL Orchestrator, which lists 17.5.0 rather than 17.5.0 to 17.5.1, BIG-IQ Centralized Management 8.4.0 to 8.4.2 on the SOAP article, and, on the five iRule articles, BIG-IP CNF and BIG-IP Next for Kubernetes 2.4.0 to 2.4.2, 2.3.0 to 2.3.4 and 2.2.0 to 2.2.5, fixed by the hotfixes 2.4.2-EHF13, 2.3.4-EHF13 and 2.2.5-EHF13 as printed. (On the per article pages the hotfix build names carry a trailing 3, such as ENG3. This site reads that digit as a footnote marker, because the hotfix note prints the same builds ending in ENG; the CNF names are copied as printed and this site did not verify whether the digit is a marker there too.) The table gives the scores, the fix and the mitigation as printed for each article. F5 evaluates only versions that have not reached End of Technical Support, and no 16.x or 15.x row appears.
Scores, fix and mitigation for each of the twelve F5 articles, as read on 20 November 2026. CVSS 3.1 first, then 4.0. Hotfix builds are those in K000164239; mitigation wording shortened.
| Article, CVE and component | Scores and fix | Mitigation as printed |
|---|---|---|
| K000163929, CVE-2026-120718, iControl REST | High 8.1; Critical 9.2. Release: 21.1.0.2, 17.5.1.10, 17.1.3.6. | Modify scripts that create REST tokens to delete them at the end of their useful life. |
| K000163936, CVE-2026-125274, iControl SOAP | High 8.8; High 8.7. Release: 21.1.0.3, 17.5.1.10, 17.1.3.6; BIG-IQ 8.4.2.3. | Only remove access for users who are not completely trusted; temporary: block iControl SOAP through self IP addresses (Port Lockdown Allow None) or the management interface. |
| K000163937, CVE-2026-123637, APM per-request policy | Critical 9.8; Critical 9.3. Hotfix only. | None. |
| K000163938, CVE-2026-125272, APM Logon Page | High 7.5; High 8.7. Hotfix only. | None. |
| K000163939, CVE-2026-120285, iRule PSC::ip_address | Critical 9.8; Critical 9.3. Hotfix only, plus CNF and Next for Kubernetes. | None. |
| K000163940, CVE-2026-120292, iRule MR::message route | Critical 9.8; Critical 9.3. Hotfix only, plus CNF and Next for Kubernetes. | None. |
| K000163941, CVE-2026-116953, iRule GENERICMESSAGE::route | Critical 9.8; Critical 9.3. Hotfix only, plus CNF and Next for Kubernetes. | None. |
| K000163942, CVE-2026-126822, APM SAML | High 7.5; High 8.7. Hotfix only. | If SAML SP is not required, remove the SAML Auth agent from the APM access policy, and restrict network access to the virtual server to trusted client networks. |
| K000163943, CVE-2026-126336, APM SAML encrypted assertions | High 7.5; High 8.7. Release: 21.1.0.3, 17.5.1.10, 17.1.3.6. | Configure high availability (HA) clustering and the HA table, to lessen the impact. |
| K000163944, CVE-2026-121812, APM and SSL Orchestrator | High 7.5; High 8.7. Hotfix only; no 17.5 line for SSL Orchestrator. | None. |
| K000163945, CVE-2026-120722, iRule DNS::scrape (BIG-IP DNS) | Critical 9.8; Critical 9.3. Hotfix only, plus CNF and Next for Kubernetes. | None. |
| K000163946, CVE-2026-121805, iRule PEM::subscriber (BIG-IP PEM) | Critical 9.8; Critical 9.3. Hotfix only, plus CNF and Next for Kubernetes. | None. |
The mitigation on the SOAP article carries a warning that matters. F5 says blocking access through the self IP addresses "prevents all access to the Configuration utility and iControl SOAP and iControl REST using the self IP address" and that "These changes may also impact other services, including breaking high availability (HA) configurations." F5 says to read its port lockdown articles first. This site adds that a change like that belongs in a maintenance window, as a suggestion, and not a quick fix.
F5 prints one command for one of the conditions. The policy trace article says the tmm.access.policytrace database key is disabled by default and that, to find out whether it is enabled, a reader enters the command below in the TMOS Shell (tmsh). It lists the key's setting and changes nothing. A reader who finds it enabled has met one of the conditions in the table above; the article lists no mitigation, so the fix is the hotfix.
list sys db tmm.access.policytrace
What the pages state, and what they do not
For each question a reader will ask, what the F5 pages and the saved BleepingComputer page print, and what none of them does.
What the F5 articles, the hotfix note and the saved BleepingComputer report state and do not state, as read on 20 November 2026.
| Question | Stated on a saved page | Not stated on any saved page |
|---|---|---|
| Which builds are fixed? | Hotfix builds per branch in nine articles; release versions in three; the build to choose in K000164239 | A regular release carrying the nine hotfix fixes, or its date; a 17.5 fix line for SSL Orchestrator |
| Is it exploited? | BleepingComputer: F5 "hasn't flagged any of these issues as exploited" | Any statement either way on an F5 page; the word exploit does not appear on them |
| What is the trigger? | "Undisclosed requests" on the ten data plane articles, with the named condition | The requests themselves; how likely code execution is |
| How many devices are affected? | BleepingComputer: "more than 177,000" BIG-IP instances and "over 14,600" with APM fingerprints (Shadowserver, not read by this site) | How many run the named policy, agent or iRule command; how many are honeypots or already patched (BleepingComputer says no information) |
| Is there a workaround? | None on eight of the twelve; four print one, each narrow | Whether the four are enough in every deployment |
| Are older branches affected? | F5 evaluates only versions not at End of Technical Support | Any 16.x or 15.x status |
| Who found them? | Eight articles credit one outside firm, two other outside parties, and two say F5 found them internally (derived) | Anything about how the flaws were reached or reported beyond that |
Where the pages differ, left unresolved
This site reports each difference as printed and resolves none.
Severity against the press word. F5: six articles Critical on CVSS 3.1, seven on 4.0, six High on 3.1 and five on 4.0. BleepingComputer: "critical" in its headline, five critical and one high. The word critical is true of half the articles on one score and of seven on the other.
Impact words. F5, article by article: "may lead to remote code execution" on six, denial of service only on four, token lifetime on one, commands by a Guest role user on one. BleepingComputer: a headline about "RCE flaws" and a sentence that attackers can "execute arbitrary code remotely" on systems with vulnerable configurations.
Fix type. F5: hotfix builds on nine, release versions on three, and a hotfix note for choosing among five builds. BleepingComputer: install the out-of-band hotfixes, and also "upgrade systems to the latest version". The F5 pages do not say that a release has the nine fixes.
Counts. The index lists twelve non-NGINX articles. BleepingComputer's report covers six of them. This site did not look for other press counts.
Product names. The index's Affected products column and the articles do not always match (derived). For K000163940, K000163941 and K000163945 the index lists BIG-IP Next for Kubernetes and not BIG-IP CNF, yet each article's own table lists BIG-IP CNF as vulnerable beside BIG-IP Next for Kubernetes, and the DNS::scrape article's Product Development line names BIG-IP CNF. A reader who used the index alone to decide whether a CNF deployment is affected would miss three articles. Also, "BIG-IP (all modules)" in the index is a product scope; the article's Description then adds the condition, such as an iRule command (this site's reading).
Update dates. The index says it was updated on 18 November. Six articles show an update on 19 November and one, K000163944, on 20 November, and none says what changed (derived from the printed dates). A reader who bookmarked an article on 18 November should read it again.
The NGINX half of the same notification
The other ten articles in the index are for NGINX. This site's earlier briefings cover that side: what nginx.org and F5 say about the NGINX 1.31.9 release, and what each page says about the NGINX2Root claims. Nothing in this briefing depends on them, and BIG-IP articles do not list NGINX as affected except to say it is not vulnerable.
What to do, in order
F5's advice, first: if you run a version in the Versions known to be vulnerable column, install a version from the Fixes introduced in column for your branch, and where the fix is an engineering hotfix, use the hotfix note to choose the build. If you run a custom engineering hotfix build, F5 says to contact F5 Support. The rest are this site's suggestions.
Take this with you
In order of value (items 2 to 7 are this site's suggestions)
- Follow F5: for each of the twelve articles that matches your modules, install the fix for your branch from its Fixes introduced in column.
- Inventory BIG-IP and BIG-IQ: versions, base releases, installed hotfixes, and which modules are provisioned (APM, DNS, PEM, CNF, SSL Orchestrator).
- Map each article to your configuration before ranking: which APM policy types, SAML service providers and iRule commands (PSC::ip_address, MR::message route, GENERICMESSAGE::route, DNS::scrape, PEM::subscriber) you actually use. Patch the matches first.
- Where the fix is a hotfix, read K000164239 before choosing the build, and plan a maintenance window: F5 says hotfixes carry no warranty or guarantee of usability.
- Check the one setting F5 gives a command for: the tmm.access.policytrace key, disabled by default.
- Treat counts of exposed BIG-IP devices as an upper bound on exposure, not a measure of risk: they count devices, not configurations.
- Keep the management interface and iControl access off the internet and limited to trusted networks, as F5's SOAP article advises for that flaw.
The question the counts leave open
If ten of the twelve articles name a specific policy, agent, iRule command or setting that has to be present, why is the number that circulates a count of devices, and which of your virtual servers actually carries one of those conditions?
Key facts
Sources
- PrimaryF5 out-of-band security notification index of 18 November 2026, saved at 12:16 UTC on 19 November 2026: the 22 articles, ten NGINX and twelve BIG-IP, BIG-IQ and related, with the affected productsF5 (MyF5 security advisory K000164058)accessed 2026-11-20
- PrimarySpecial considerations for engineering hotfixes released on 18 November 2026, read at 15:25 UTC on 20 November 2026: the five builds and which to chooseF5 (MyF5 support solution K000164239)accessed 2026-11-20
- PrimaryBIG-IP iControl REST vulnerability CVE-2026-120718, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163929)accessed 2026-11-20
- PrimaryBIG-IP and BIG-IQ iControl SOAP vulnerability CVE-2026-125274, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163936)accessed 2026-11-20
- PrimaryBIG-IP APM vulnerability CVE-2026-123637, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163937)accessed 2026-11-20
- PrimaryBIG-IP APM vulnerability CVE-2026-125272, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163938)accessed 2026-11-20
- PrimaryBIG-IP iRule vulnerability CVE-2026-120285, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163939)accessed 2026-11-20
- PrimaryBIG-IP iRule vulnerability CVE-2026-120292, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163940)accessed 2026-11-20
- PrimaryBIG-IP iRule vulnerability CVE-2026-116953, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163941)accessed 2026-11-20
- PrimaryBIG-IP APM SAML vulnerability CVE-2026-126822, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163942)accessed 2026-11-20
- PrimaryBIG-IP APM vulnerability CVE-2026-126336, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163943)accessed 2026-11-20
- PrimaryBIG-IP APM and BIG-IP SSL Orchestrator vulnerability CVE-2026-121812, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163944)accessed 2026-11-20
- PrimaryBIG-IP iRule vulnerability CVE-2026-120722, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163945)accessed 2026-11-20
- PrimaryBIG-IP iRule vulnerability CVE-2026-121805, read at 15:23 to 15:25 UTC on 20 November 2026: description, impact, the table of versions, fixes, severity and CVSS 3.1 and 4.0, and mitigationF5 (MyF5 security advisory K000163946)accessed 2026-11-20
- Reported byF5 releases emergency patches for critical BIG-IP RCE flaws, 20 November 2026, saved at about 15:18 UTC on 20 November 2026: the press wording compared with the F5 articles; its Shadowserver counts are relayed, not read by this siteBleepingComputeraccessed 2026-11-20


