BigFix RunBookAI CVE-2025-31951

Requires valid user credentials with authenticated access to HCL BigFix RunBookAI 11.2 (PR:L - low-privileged authentication). The vulnerability affects the specific component handling command inputs within RunBookAI, which is part of the standard automation workflow functionality. Exploitation requires the attacker to interact with command input fields or API endpoints that feed into the vulnerable component's processing logic. No special deployment configuration or non-default feature enablement is indicated in the available data. Exploitation is limited by the authentication requirement, meaning attackers must first obtain legitimate BigFix RunBookAI user credentials through phishing, credential compromise, or insider access. The network attack vector (AV:N) allows remote exploitation once authentication is achieved.

This represents a high-severity threat with meaningful real-world risk for organizations using BigFix RunBookAI. The CVSS 8.8 score reflects critical impact dimensions: network attack vector (AV:N) enabling remote exploitation, low attack complexity (AC:L) requiring no special conditions, and complete compromise of confidentiality, integrity, and availability (C:H/I:H/A:H). The requirement for low-level authentication (PR:L) provides minimal defense as RunBookAI users typically have legitimate access credentials for automation tasks. No EPSS score, KEV listing, or public exploit code is currently identified in available data, suggesting exploitation remains theoretical rather than widespread. However, command injection vulnerabilities in automation platforms are highly attractive targets because successful exploitation grants attacker commands the same elevated privileges the automation system uses to manage infrastructure. The SSVC framework would likely classify this as exploitable with significant technical impact given the automation context. Organizations running version 11.2 should prioritize remediation based on the credential access controls in their environment and whether the vulnerable component is exposed to broader user populations.

An authenticated attacker with low-privilege RunBookAI credentials crafts a malicious automation request containing injected shell metacharacters (such as semicolons, pipes, or command substitution syntax) in input fields processed by the vulnerable component. The unvalidated input is passed to an underlying command interpreter, allowing the attacker to append arbitrary system commands that execute with the privileges of the RunBookAI service, potentially gaining control over the automation server and pivoting to managed endpoints throughout the BigFix infrastructure.