CVE-2024-24590
📋 TL;DR
This vulnerability in Allegro AI's ClearML client SDK allows remote code execution through deserialization of untrusted data. An attacker can upload a malicious artifact that executes arbitrary code on a user's system when interacted with. Users of ClearML versions 0.17.0 through 1.14.2 are affected.
💻 Affected Systems
- Allegro AI ClearML Client SDK
📦 What is this software?
Clearml by Clear
⚠️ Risk & Real-World Impact
Worst Case
Complete compromise of the user's system leading to data theft, ransomware deployment, or lateral movement within the network.
Likely Case
Execution of malicious code on individual user workstations, potentially stealing credentials or installing malware.
If Mitigated
Limited impact if proper network segmentation and least privilege principles are enforced.
🎯 Exploit Status
Exploitation requires uploading a malicious artifact and user interaction, but the deserialization vulnerability itself is straightforward to exploit.
🛠️ Fix & Mitigation
✅ Official Fix
Patch Version: 1.14.3
Vendor Advisory: https://github.com/allegroai/clearml/releases/tag/1.14.3
Restart Required: No
Instructions:
1. Update ClearML client SDK to version 1.14.3 or later. 2. Run: pip install --upgrade clearml>=1.14.3
🔧 Temporary Workarounds
Restrict artifact uploads
allLimit who can upload artifacts to trusted users only and implement file type restrictions.
Network segmentation
allIsolate ClearML instances from critical systems and implement strict outbound firewall rules.
🧯 If You Can't Patch
- Implement strict access controls on artifact upload functionality
- Monitor for suspicious artifact uploads and user interactions with artifacts
🔍 How to Verify
Check if Vulnerable:
Check ClearML version: python -c "import clearml; print(clearml.__version__)"Check Version:
python -c "import clearml; print(clearml.__version__)"Verify Fix Applied:
Verify version is 1.14.3 or higher using the same command📡 Detection & Monitoring
Log Indicators:
- Unusual artifact uploads from unexpected sources
- Process execution from ClearML artifact directories
Network Indicators:
- Outbound connections from ClearML processes to suspicious IPs
SIEM Query:
Process creation where parent process contains 'clearml' AND command line contains suspicious patterns