| Description |
FortiGuard Labs have observed exploitation attempts of CVE-2025-3248 in Langflow.
Langflow is a Python-based application that allows users to visually build AI agents and workflows.
CVE-2025-3248 is an authentication bypass flaw that enables unauthenticated attackers to remotely execute arbitrary Python code by sending a crafted HTTP request to the vulnerable endpoint.
The following version of Langflow are vulnerable to CVE-2025-3248:
|
||||||
|
CVE ID |
CVE-2025-3248 (https://nvd.nist.gov/vuln/detail/CVE-2025-3248) | ||||||
| NDR Cloud Detection Rule |
FortiNDR Cloud v25.2b+
|
||||||
|
Playbook |
N/A |
||||||
| Threat Hunting | FortiNDR Cloud users can use the following IOCs from Fortinet to hunt for “Langflow Unauth RCE Attack” related activities. IOC source: https://www.fortiguard.com/outbreak-ioc?tag=Langflow%20Unauth%20RCE All IOCs listed above have been added to Threat Intelligence Intel. |
||||||
| Suricata Coverage |
Customers can create custom investigation/detections using the Suricata signatures below: 2061448 -> ET WEB_SPECIFIC_APPS Langflow AI Unauthenticated Remote Code Execution via Code Validation Endpoint (CVE-2025-3248) |
||||||
| Other Fortinet Products | For more details regarding mitigating the vulnerability by utilizing Fortinet products, please refer to: https://www.fortiguard.com/outbreak-alert/langflow-unauth-rce |
The Fortinet Security Fabric brings together the concepts of convergence and consolidation to provide comprehensive cybersecurity protection for all users, devices, and applications and across all network edges.
Copyright 2025 Fortinet, Inc. All Rights Reserved.