Improper Access Control in IBM Langflow OSS Leads to Remote Code Execution

IBM Langflow OSS, an open-source visual framework for building multi-agent AI applications, is affected by a critical vulnerability, CVE-2026-104335. This flaw exists in versions 1.0.0 through 1.12.2 and stems from improper access control mechanisms within the application. An attacker with authenticated access can leverage this vulnerability to execute arbitrary code on the underlying host system. Given that Langflow environments often handle sensitive API keys, model configurations, and data processing tasks, this vulnerability represents a significant risk for unauthorized access, lateral movement, and data exfiltration. Organizations deploying IBM Langflow OSS should prioritize upgrading to a patched version once available and enforce strict access controls on the management interface.

Impact

Successful exploitation of this vulnerability allows an authenticated attacker to bypass intended access restrictions and achieve arbitrary code execution. This could result in a complete compromise of the Langflow application instance, unauthorized access to connected AI services or databases, and the potential for exfiltration of intellectual property or sensitive model data.

Recommendation

Prioritize the upgrade of all IBM Langflow OSS instances to a version later than 1.12.2. Monitor administrative access logs for unusual login patterns or actions performed by users with elevated privileges. Given the lack of specific exploitation telemetry, focus on identifying unauthorized attempts to access or modify workflow configurations within the Langflow interface.