CVE-2026-126336: F5 BIG-IP APM: SAML encrypted assertion TMM DoS
HIGHA remote, unauthenticated attacker can crash the Traffic Management Microkernel (TMM) on F5 BIG-IP systems running APM with a SAML Service Provider agent that has encrypted assertions enabled, which disrupts traffic while TMM restarts. This is a data plane denial of service with no control plane exposure and no confidentiality or integrity impact (CVSS 7.5, CWE-191 integer underflow). It is not in CISA KEV, no public exploit or Nuclei template is known, and EPSS data is not yet available, so near-term mass exploitation looks unlikely but is not ruled out for an internet-facing, low-complexity network bug. BIG-IP is not an AI package. It matters to AI programs only where it fronts AI/ML portals, inference APIs or SSO gateways, where an outage would cut access to those services. Check whether APM with SAML SP encrypted assertions is in use, apply the fixed release listed in F5 advisory K000163943, and note that versions past End of Technical Support are not evaluated, so they need an upgrade path.
What is the risk?
Overall risk is moderate to high for exposed BIG-IP deployments that match the configuration, and low for everything else. Exploitation needs only network access, with low complexity, no privileges and no user interaction. The impact is limited to availability (A:H). The vulnerable condition is narrow: APM provisioned, a virtual server with an APM Access profile, and a SAML SP agent with encrypted assertions enabled. There is no known active exploitation, no KEV listing, no public exploit and no EPSS score (N/A). Package-level AI signals are absent (risk score 0/100, no dependents data). The classification as an AI tool appears to be a catalog artifact, and the real AI relevance is indirect, through the services BIG-IP fronts.
How does the attack unfold?
How severe is it?
What is the attack surface?
What should I do?
1 step1) Identify BIG-IP devices with APM provisioned and virtual servers using an Access profile with a SAML SP agent and encrypted assertions enabled. 2) Upgrade to the fixed software version listed in F5 advisory K000163943. Versions at End of Technical Support are not evaluated, so plan an upgrade for those. 3) Until patched, review whether encrypted assertions are required on the SAML SP agent and follow any mitigation or workaround in the advisory. Limit exposure of the affected virtual server to trusted networks or IdP source ranges where feasible. 4) Detection: monitor for unexpected TMM restarts, core files, and HA failovers, and alert on repeated tmm crash messages in /var/log/ltm. Watch for bursts of unusual requests to SAML SP endpoints on the virtual server. 5) Confirm HA failover behaviour and keep a tested runbook for TMM crash recovery.
How is it classified?
Which compliance frameworks are affected?
This CVE is relevant to:
Frequently Asked Questions
What is CVE-2026-126336?
A remote, unauthenticated attacker can crash the Traffic Management Microkernel (TMM) on F5 BIG-IP systems running APM with a SAML Service Provider agent that has encrypted assertions enabled, which disrupts traffic while TMM restarts. This is a data plane denial of service with no control plane exposure and no confidentiality or integrity impact (CVSS 7.5, CWE-191 integer underflow). It is not in CISA KEV, no public exploit or Nuclei template is known, and EPSS data is not yet available, so near-term mass exploitation looks unlikely but is not ruled out for an internet-facing, low-complexity network bug. BIG-IP is not an AI package. It matters to AI programs only where it fronts AI/ML portals, inference APIs or SSO gateways, where an outage would cut access to those services. Check whether APM with SAML SP encrypted assertions is in use, apply the fixed release listed in F5 advisory K000163943, and note that versions past End of Technical Support are not evaluated, so they need an upgrade path.
Is CVE-2026-126336 actively exploited?
No confirmed active exploitation of CVE-2026-126336 has been reported, but organizations should still patch proactively.
How to fix CVE-2026-126336?
1) Identify BIG-IP devices with APM provisioned and virtual servers using an Access profile with a SAML SP agent and encrypted assertions enabled. 2) Upgrade to the fixed software version listed in F5 advisory K000163943. Versions at End of Technical Support are not evaluated, so plan an upgrade for those. 3) Until patched, review whether encrypted assertions are required on the SAML SP agent and follow any mitigation or workaround in the advisory. Limit exposure of the affected virtual server to trusted networks or IdP source ranges where feasible. 4) Detection: monitor for unexpected TMM restarts, core files, and HA failovers, and alert on repeated tmm crash messages in /var/log/ltm. Watch for bursts of unusual requests to SAML SP endpoints on the virtual server. 5) Confirm HA failover behaviour and keep a tested runbook for TMM crash recovery.
What systems are affected by CVE-2026-126336?
This vulnerability affects the following AI/ML architecture patterns: Model serving behind BIG-IP gateways, SAML SSO access to ML platforms, Inference API load balancing, Internal AI portals and notebooks.
What is the CVSS score for CVE-2026-126336?
CVE-2026-126336 has a CVSS v3.1 base score of 7.5 (HIGH).
What is the AI security impact?
Affected AI Architectures
MITRE ATLAS Techniques
AML.T0029Denial of AI ServiceAML.T0049Exploit Public-Facing ApplicationCompliance Controls Affected
What are the technical details?
Original Advisory
Description: When BIG-IP APM is provisioned and a virtual server is configured with an APM Access profile and a SAML Service Provider (SP) agent with encrypted assertions enabled, undisclosed requests can cause the Traffic Management Microkernel (TMM) to terminate. Impact: 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. There is no control plane exposure; this is a data plane issue only. Note: Software versions which have reached End of Technical Support (EoTS) are not evaluated.
Exploitation Scenario
An attacker scans the internet for BIG-IP virtual servers exposing an APM SAML SP endpoint, for example the SSO gateway in front of a company's MLflow, Jupyter or internal LLM portal. They send the undisclosed crafted requests to the SAML SP agent that processes encrypted assertions, triggering an integer underflow that terminates TMM. By repeating the requests, they keep the data plane cycling, cutting access to the AI services behind the gateway and disrupting data scientists, inference clients and automated pipelines that depend on that path. No credentials are needed.
Weaknesses (CWE)
CWE-191 — Integer Underflow (Wrap or Wraparound): The product subtracts one value from another, such that the result is less than the minimum allowable integer value, which produces a value that is not equal to the correct result.
Source: MITRE CWE corpus.
CVSS Vector
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:HReferences
Timeline
Related Vulnerabilities
CVE-2026-3366010.0TensorFlow: type confusion NPD in tensor conversion
Same attack type: DoSCVE-2023-256689.8TensorFlow: unauthenticated RCE via heap buffer overflow
Same attack type: DoSCVE-2022-235879.8TensorFlow: integer overflow in Grappler enables RCE
Same attack type: DoSCVE-2022-359399.8TensorFlow: ScatterNd OOB write enables RCE/crash
Same attack type: DoSCVE-2022-419009.8TensorFlow: heap OOB RCE in FractionalMaxPool op
Same attack type: DoS