The sanitize_jinja() function in the automation-controller uses two regular expressions to reject user-supplied Jinja templates, but these patterns stop at the first interior '}' or '%' character. This allows Jinja expressions containing inner braces (e.g., empty dictionaries) to pass validation while remaining valid. Since sanitize_jinja() is the sole guard on several launch-time fields such as ad-hoc command module_args, Machine-credential username, become_method, become_user, and inventory host names, a low-privileged user can inject Jinja that ansible-core evaluates. This leads to arbitrary command execution bypassing administrator allowlists and disclosure of secrets from credentials the attacker should not access.

Exploitation of this vulnerability allows an attacker with low privileges to execute arbitrary commands within the execution environment, bypassing administrative controls such as AD_HOC_COMMANDS module allowlists. Additionally, it enables unauthorized disclosure of secrets from credentials that the attacker cannot normally read by templating environment variables injected by co-attached credentials. This compromises confidentiality and integrity of the automation environment.

A security update is available from Red Hat for Ansible Automation Platform 2.6 that addresses this vulnerability. Users should apply the official patch as provided in Red Hat Advisory RHSA-2026:71113 to remediate this issue. No additional mitigation steps are required beyond applying the vendor's fix.