Python AI library hacked to spread malware

A popular AI model has fallen victim to a hacker attack: unknown people compromised the Python library Ultralytics YOLO11 and used it to spread malware. The incident shows the vulnerability of open source projects and calls for vigilance.

Attack on AI library: Malware instead of object recognition

This time it hit the popular Python library Ultralytics YOLO11, an AI model for computer vision and object recognition. Hackers compromised two versions of the software and foisted a cryptominer on users.

The malware, known as XMRig, uses the computing power of infected devices to mine the cryptocurrency Monero. According to initial findings, versions 8.3.41 and 8.3.42, which were published via the Python Package Index (PyPI) platform, are affected.

Ultralytics confirms the incident

How The Hacker News reports Glenn Jocher, the founder of Ultralytics, already confirmed the incident and promised comprehensive clarification. Jocher explained:

We confirm that Ultralytics versions 8.3.41 and 8.3.42 were compromised by a malicious code injection targeting cryptocurrency mining. Both versions were immediately removed from PyPI. Glenn Jocher, founder of Ultralytics

The compromised versions were therefore removed from the PyPI repository as the problem was identified and a cleaned version 8.3.43 was released at the same time. YOLO11 is an extremely popular open source project. It receives hundreds of thousands of downloads every day. This popularity makes it an attractive target for cybercriminals.

Clever approach of the hackers

Security researchers suspect that the attackers exploited a known vulnerability in GitHub Actions. The attackers did this in a sophisticated way: They did not manipulate the source code on GitHub, but rather the build environment for the PyPI deployment. This allowed them to make unnoticed changes after the actual code review.

Recommendations for users

Following the incident, Ultralytics plans to implement sandboxing in the desktop application in the future to better protect us against these types of attacks. However, library users are currently advised to exercise extreme caution. You should take the following steps:

• Check which version you are using
• If necessary, update to the latest, cleaned version
• Also check indirect use via dependencies in other projects
• Conduct a thorough check of all systems for possible infections

This attack joins a series of similar incidents in which popular open source projects were exploited to distribute malware. It highlights the need to rethink and improve the security of the entire software supply chain.