Threat database

CVE-2025-64703 is a medium-severity vulnerability in the open-source AI assistant MaxKB (versions prior to 2.3.1) developed by 1Panel-dev. The flaw allows users with limited privileges to extract sensitive information via Python code execution within a sandboxed tool module. Although the process runs in a sandbox, the isolation is insufficient to prevent unauthorized data exposure. The vulnerability impacts confidentiality, integrity, and availability to a limited extent and does not require user interaction but does require some level of privilege. The issue was addressed in version 2.3.1. European organizations using vulnerable MaxKB versions may face risks of sensitive data leakage, potentially affecting enterprise AI deployments.

MaxKB is an open-source AI assistant for enterprise. Prior to version 2.0.0, the sandbox design rules can be bypassed because MaxKB only restricts the execution permissions of files in a specific directory. Therefore, an attacker can use the `shutil.copy2` method in Python to copy the command they want to execute to the executable directory. This bypasses directory restrictions and reverse shell. Version 2.0.0 fixes the issue.