F5 has rolled out emergency patches for six severe vulnerabilities in its BIG-IP product line, flaws that hackers could exploit to knock websites offline or run malicious code—without needing to authenticate. These issues stem from memory corruption in the device’s traffic-handling component, jeopardizing services even if the admin interface remains secure.
What’s at Risk?
BIG-IP appliances act as gatekeepers between users and the servers hosting a website. The Traffic Management Microkernel (TMM) module manages client connections. If TMM crashes due to an attack, web services and remote access portals relying on BIG-IP will go down—even though the underlying servers are still operating. The danger lies in the data plane, which processes live traffic. It doesn’t matter if the management interface isn’t exposed; a vulnerable virtual server that’s reachable from the internet could still be exploited.
Details on the Vulnerabilities and Affected Versions
Five of the six critical bugs are linked to iRules, a scripting framework used to customize how BIG-IP handles incoming traffic—these are identified as CVE-2026-120285, 120292, 116953, 120722, and 121805. The sixth, CVE-2026-123637, targets the Access Policy Manager when it’s using a per-request policy with both the Client Information and Endpoint Security modules. While there’s another serious flaw, CVE-2026-126822, that also causes outages without login, it’s ranked below “critical” and sits outside this emergency update.
The vulnerabilities span several BIG-IP product lines including those for Kubernetes and cloud-native network functions. Specifically, affected versions include 17.1.0-17.1.3, 17.5.0-17.5.1, and version 21.1.0, along with some builds of Next for Kubernetes and CNF. F5 has published patched builds for each impacted version—meaning administrators must check their current release or engineering build to apply the correct hotfix.
Urgent Mitigation Steps
F5 doesn’t list any workarounds for these flaws. Organizations should focus first on virtual servers that are internet-facing or handle critical services. Before applying any hotfix, it’s essential to record the existing system build and test updates in non-production environments to avoid unintended disruptions. Once patched, it’s important to verify that websites and remote-access portals are genuinely reachable—logging into the management interface alone doesn’t guarantee live traffic is successfully passing through.
No evidence of these vulnerabilities being exploited has surfaced so far. That sets this situation apart from the BIG-IP OAuth zero-day disclosed in September, which was already in use by attackers at the time.
These fixes cover BIG-IP TMOS releases 17.1.x, 17.5.x, 21.1.x, as well as specific versions of BIG-IP Next for Kubernetes and Cloud-Native Network Functions. For teams running custom builds, F5 advises reaching out to support for bespoke patches.
Analytically, this incident underscores how even mature, widely deployed systems like BIG-IP can harbor critical flaws that bridge the gap between denial-of-service and remote code execution. As enterprises increasingly depend on edge and traffic-management devices, ensuring rapid patch management is no longer optional—it’s business continuity. Organizations should invest in monitoring traffic process health and build governance, not just interface access. What to watch for next: whether any of the patched builds themselves introduce new issues in real-world deployments, and how attackers leverage these memory corruption vectors in future campaigns.