Langflow's Critical Authenticated RCE Flaw Targeted in 1,423 Honeypot Exploit Attempts

A critical vulnerability in Langflow (CVE-2026-105740) allows any authenticated user to execute arbitrary OS commands by supplying a malicious, unvalidated input in the MCP server command field — a fundamental failure of input validation and secure design. With over 1,400 exploit attempts recorded against honeypots, this flaw is already in active threat actor playbooks, meaning unpatched instances face imminent compromise. The risk is compounded by the fact that only authenticated access is required, meaning any compromised or low-privilege account can escalate to full OS-level control. This incident underscores how quickly critical RCE vulnerabilities in popular open-source AI tools become weaponized, often before organizations have completed their patch cycles. Organizations running AI/ML platforms must treat them with the same security rigor as production application servers.