Cairn AI Agents Found MONAI Label Zero-Day and Gained Server Root Access

Summary
Amazon says Cairn AI agents exploited a path traversal zero-day in MONAI Label to write files and gain root access. The campaign targeted mainly East Asian educational institutions; maintainers released a fix on November 18.
Key points
- Amazon Threat Intelligence says a suspected Chinese-linked actor used Cairn between September 8 and 10, 2026, mainly targeting educational institutions in East Asia outside China.
- The agents reportedly found a path traversal zero-day in MONAI Label, enabling arbitrary file writes.
- In one session, agents wrote a Python reverse shell to sitecustomize.py and triggered an API endpoint to gain root access.
- Amazon said it reported the flaw to MONAI Label maintainers, who released a security update on November 18.
- Amazon attributed the campaign to a Chinese-nexus actor with moderate confidence, citing retrieved prompt language, activity timestamps, and target selection.
- Other reported attempts failed, including a 16-hour Redis intrusion effort and more than 30,000 credential guesses against a university GPU server.
Article Details
- Event Type
- Reported exploitation of a path traversal zero-day in MONAI Label using the Cairn autonomous AI agent platform.
- Impact
- The flaw allowed attackers to write files to arbitrary server locations. Amazon reported that the agents used it to establish root access to a server; maintainers released a security update on November 18, 2026.
Indicators of compromise
| Type | Indicator | Context |
|---|---|---|
| IPV4 | 142[.]4[.]107[.]34 | IP address Amazon shared as associated with the campaign, observed September 8–10, 2026. |
| IPV4 | 142[.]4[.]98[.]145 | IP address Amazon shared as associated with the campaign, observed September 8–10, 2026. |
