
Chinese-Linked Actors Use Cairn AI Platform to Discover and Exploit MONAI Label Zero-Day
Hackers leveraged the autonomous AI agent platform Cairn to identify a path traversal zero-day in the open-source MONAI Label medical imaging application and chain it into full root compromise of a target server. The suspected Chinese-nexus operators simply instructed the agents to obtain shell access, after which the AI independently mapped the API, located the flaw, and wrote a reverse shell into sitecustomize.py within eleven minutes. Amazon Threat Intelligence discovered the campaign after the attackers left their Cairn instance publicly exposed without authentication. The activity, observed between 8 and 10 September 2026, primarily targeted East Asian educational institutions and also included failed attempts against Redis and GPU servers. Amazon responsibly disclosed the vulnerability, which MONAI Label maintainers patched on 18 November 2026. Researchers assess the operator with moderate confidence as Chinese-linked based on prompt language, timestamps, and target selection.
BleepingComputerNov 20, 2026Translated from English