The vulnerability arises from incomplete input validation in the sanitize_jinja() function of the automation-controller component in Red Hat Ansible Automation Platform 2. The function uses two regular expressions to reject user-supplied Jinja templates but stops at the first interior '}' or '%' character, allowing nested braces (e.g., empty dicts) to pass through. Since sanitize_jinja() is the only guard on several critical fields (ad-hoc command module_args, machine credential username, become_method, become_user, and inventory host names), a low-privileged user can inject Jinja code that ansible-core evaluates in the execution environment. This bypasses administrator allowlists for ad-hoc commands and enables execution of arbitrary commands and disclosure of secrets from credentials the attacker should not access. The CVSS v3.1 score is 7.1 (High) with network attack vector, low complexity, low privileges required, no user interaction, unchanged scope, high confidentiality impact, low integrity impact, and no availability impact. No known exploits in the wild have been reported. The vendor advisory is available at https://access.redhat.com/security/cve/CVE-2026-84714.

An attacker with low privileges can inject malicious Jinja templates into several launch-time fields that are evaluated by ansible-core, enabling arbitrary command execution within the execution environment. This bypasses administrative restrictions such as AD_HOC_COMMANDS module allowlists. Additionally, the attacker can disclose secrets from credentials they are not authorized to read by templating environment variables injected by co-attached credentials. The vulnerability compromises confidentiality significantly and integrity to a lesser extent, without affecting availability.

The vendor advisory does not explicitly state that a fix is currently available or provide patch links. Patch status is not yet confirmed — check the Red Hat advisory at https://access.redhat.com/security/cve/CVE-2026-84714 for current remediation guidance. Until a fix is released, administrators should consider restricting access to affected fields and carefully review user privileges to limit exposure. Monitoring vendor updates for patches or official mitigations is recommended.