CVE-2024-10950
📋 TL;DR
This vulnerability allows attackers to execute arbitrary code on servers running vulnerable versions of binary-husky/gpt_academic through prompt injection in the CodeInterpreter plugin. Users running version 3.83 or earlier are affected, potentially leading to complete server compromise.
💻 Affected Systems
- binary-husky/gpt_academic
📦 What is this software?
Gpt Academic by Binary Husky
⚠️ Risk & Real-World Impact
Worst Case
Full remote code execution leading to complete server takeover, data exfiltration, lateral movement, and persistent backdoor installation.
Likely Case
Attacker gains shell access to the application server, allowing file system access, credential theft, and further exploitation of the environment.
If Mitigated
Limited impact with proper network segmentation and least privilege controls, potentially containing damage to isolated application components.
🎯 Exploit Status
Requires ability to submit prompts to the CodeInterpreter plugin. No public exploit code available as of analysis.
🛠️ Fix & Mitigation
✅ Official Fix
Patch Version: > 3.83
Vendor Advisory: https://huntr.com/bounties/9abb1617-0c1d-42c7-a647-d9d2b39c6866
Restart Required: No
Instructions:
1. Update to latest version of gpt_academic (post 3.83). 2. Verify CodeInterpreter plugin has proper sandboxing implemented. 3. Test functionality after update.
🔧 Temporary Workarounds
Disable CodeInterpreter Plugin
allTemporarily disable the vulnerable plugin until patching is possible.
Modify configuration to disable CodeInterpreter plugin
Implement Input Validation
allAdd strict input validation and sanitization for prompts before processing.
Implement prompt validation rules in application code
🧯 If You Can't Patch
- Implement network segmentation to isolate the vulnerable application from critical systems
- Deploy application-level firewall with rules to detect and block suspicious prompt patterns
🔍 How to Verify
Check if Vulnerable:
Check gpt_academic version and verify if CodeInterpreter plugin is enabled in configuration.Check Version:
Check package version or application metadata for version numberVerify Fix Applied:
Confirm version is > 3.83 and test CodeInterpreter functionality with safe prompts.📡 Detection & Monitoring
Log Indicators:
- Unusual prompt patterns in CodeInterpreter logs
- Unexpected process execution from application context
- Error messages related to code execution failures
Network Indicators:
- Outbound connections from application server to unexpected destinations
- Unusual network traffic patterns from application
SIEM Query:
source="gpt_academic" AND (event="code_execution" OR event="plugin_error")