PANDASAI · CVE-2024-12366 · AI agents & MCP
NO FIX YET · SEE WHAT TO DO
PandasAI's interactive prompt function can be tricked by prompt injection into running arbitrary Python code on the host. An attacker who controls prompt input can achieve remote code execution and pivot to connected services.
first public Feb 11, 2025 · 2 sources
Check it with your coding agent
It reads this page and tells you if your code or app is affected.
Teams that run PandasAI (pypi package pandasai, versions 2.4.2 and earlier, per the product data) and expose its interactive prompt function to untrusted users. Applications that use PandasAI only with trusted inputs and no user-facing prompt are less exposed. The sources do not say whether the flaw depends on a specific configuration.
Not affected if no-user-prompt-input
Who cares data-analysis, llm, pandasai, python
pip show pandasai; versions 2.4.2 and earlier are listed as affected.pandasai (pypi): fixed version unknown; sources list affected versions up to 2.4.2 and state no fixed version in the product data.PandasAI uses an interactive prompt function that is vulnerable to prompt injection and run arbitrary Python code that can lead to Remote Code Execution (RCE) instead of the intended explanation of the natural language processing by the LLM. The security controls of PandasAI (2.4.3 and earlier) fail to distinguish between legitimate and malicious inputs, allowing the attackers to manipulate the system into executing untrusted code, leading to untrusted code execution (RCE), system compromise, or pivoting attacks on connected services.
This page as .md · .json · CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
GitHub Advisory Database, CC-BY 4.0 · NVD data is provided by NIST and is public domain. Posts quoted belong to their authors.
Are you affected?
Ask from here. InfraStack checks the affected versions against your stack and answers in your inbox.
InfraStack checks the affected versions against your stack and answers in Slack, with the evidence and the fix.
InfraStack checks it against your stack and answers on your report, with the evidence and the fix.