Database/Control plane, storage & DevOps
Ansible: internal templating can drop the unsafe marker, allowing template injection from supplied data
Impact
Ansible marks externally sourced data as unsafe so it is never re-templated. Internal templating operations could remove that designation, so a specially crafted file consumed by a play gets evaluated as a Jinja template on the controller. Anything that can place or influence such a file - a fact file, inventory data, a registered result from a managed host - can then run expressions in the controller's context, and the controller is usually the host holding fleet SSH keys and vault secrets. For an operator who drives node provisioning, driver rollout or BMC configuration from Ansible, controller compromise is fleet compromise.
Who can reach it
An actor able to supply or modify data that a play templates - a file read into a variable, inventory or facts from a managed host. No access to the controller itself is required, but the play must consume the attacker-influenced data. The flaw executes on the controller, not the managed node.
What to do
Patch ansible-core from your distribution or the Red Hat errata (RHSA-2023:7773 for Ansible Automation Platform 2.4); ansible-core 2.16.0 carries the fix. The controller is a single host and Ansible runs per-invocation, so there is no node drain or reboot - update the package and let the next run use it. Nothing to change on managed nodes.
References
Related entries
- OpenVPN: Stack overflow in the interactive serviceCVE-2024-27459 · OpenVPNHigh
- Intel QuickAssist Technology (QAT) software and driversCVE-2024-31858 · Intel QuickAssist Technology (QAT) software and drivers - QAT software before 2.2.0, with a 2025 batch through 2.6.0High
- IBM Storage Scale GUI (local privilege escalation): A local privilege escalation in the Storage Scale GUI availableCVE-2024-31891 · IBM Storage Scale GUI (local privilege escalation)High
- Linux HID/amd_sfh - driver_data freed after HID device destruction: A use-after-free in the AMD Sensor Fusion Hub HIDCVE-2024-46746 · Linux HID/amd_sfh - driver_data freed after HID device destructionHigh
- N-able N-central: Deserialization of untrusted data allowing local code execution on the RMM serverCVE-2025-8875 · N-able N-centralHigh
- ansible-core: malicious Galaxy role injects git flags to run code on the machine installing itCVE-2026-11332 · ansible-core (ansible-galaxy role install, git argument injection via meta/requirements.yml)High
This entry is curated: imported from vendor advisories with machine assistance, not yet individually verified. Confirm against your vendor's advisory before acting, and report anything wrong.