Hackers use AI tool Cairn to find a zero-day and take a server
Amazon Threat Intelligence says a Chinese-nexus actor used the Cairn AI agent platform to find a zero-day in MONAI Label and gain root on a server. The maintainers released a fix on November 18.
By JetherVerse News

Amazon Threat Intelligence reported on November 20, 2026 an attack in which hackers used an autonomous AI agent platform called Cairn to find a zero-day flaw in a medical imaging application and take control of a server. BleepingComputer covered the findings.
The attackers told the agents to obtain shell access and left the rest to the software. Amazon says a Chinese-nexus threat actor ran the campaign between September 8 and 10, 2026, aimed mostly at East Asian educational institutions outside China. Amazon found the activity because the operators left their Cairn instance open to the public without authentication. Cairn is a legitimate open-source penetration-testing framework built for research and defense. The tool coordinates several AI agents in a cyber operation. Gambit Security earlier documented abuse of Cairn in a credit card theft operation in September, and Google reported its use against a US aerospace organization.
The agents did not always win. A 16-hour effort of 148 actions against a Redis server failed. Over six hours, the agents tried more than 30,000 credential combinations on a university GPU server and got nowhere. The breakthrough came against MONAI Label, an open-source application where AI helps researchers label medical images. The agents found a path traversal flaw, which allows an attacker to write files to any location on the server. They went from mapping the API to exploiting the flaw in 11 minutes. In a later session they wrote a Python reverse shell into the sitecustomize.py file, which loads when Python starts, then triggered an API endpoint to gain root access. The operator then gave the agents a proxy into a second university network running the same software.
Amazon reported the flaw to the MONAI Label maintainers, who released a fix on November 18. Amazon attributes the campaign to a Chinese-nexus actor with moderate confidence, based on prompt language, activity timestamps and target choice. The company shared two IP addresses tied to the campaign. Amazon warns AI agents find and exploit flaws faster than security teams react, and urges more automated defense in place of manual investigation.
If you run MONAI Label, update to the November 18 release now. If you expose any tool to the internet, require authentication first.
Read the full report at BleepingComputer.
Source: BleepingComputer
Get the news in your inbox
The tech stories that matter from JetherVerse News, with the source on every one.

