Amazon Threat Intelligence attributes, with moderate confidence, a campaign to a Chinese-nexus actor who used the Cairn pentesting framework to coordinate AI agents against East Asian university servers in September. The agents moved from mapping the API to exploiting a path traversal flaw in MONAI Label within 11 minutes and gained root, partly by writing a reverse shell to sitecustomize.py. The maintainers released a security update on 18 November, so teams running MONAI Label should apply it.
