Skip to main content
Fundy Tech Solutions Inc.
All news
Security AlertCritical

AI agents found a MONAI Label zero-day and got root access

Nov 21, 2026 · Summarized by Fundy Tech from 2 sources

The 30-second version

Amazon says autonomous AI agents used by a China-linked threat actor found a zero-day in MONAI Label and reached root on a university server. The flaw affects version 0.8.5 and earlier, and no fixed version is available yet.

Key points

  • Amazon linked the campaign to university servers and an exposed Cairn control panel.
  • The flaw lets unauthenticated users write files outside MONAI Label storage folders.
  • Amazon said the agents reached file-writing in about 11 minutes.
  • Maintainers merged a source fix on October 19, but no installable patch exists yet.

Why it matters

If your organization runs MONAI Label, the default setup can be exposed without a login. That can let an attacker write files where they should not, which may open the door to code execution depending on how the server is configured. For small teams and households, the main lesson is to patch and review any internet-facing systems quickly. If you use research or imaging software, check whether authentication is turned on and whether a fixed release has been issued.

Who’s affected: MONAI Label 0.8.5 and earlier

What to do

  • Update MONAI Label as soon as a fixed release is available.
  • Turn on authentication on any exposed server.
  • Restrict network access to medical imaging systems.
  • Review logs for unusual file writes or Python startup files.

Original reporting

Share

Not sure if this affects you?

Fundy Tech helps businesses and households across Atlantic Canada stay patched, protected and up to speed. Tell us what you’re running and we’ll tell you straight whether you need to act.

More security alerts