Amazon: Suspected Chinese hackers used AI tool Cairn to find MONAI Label zero-day and gain root
Amazon Threat Intelligence found the campaign because the attackers left their Cairn instance open without authentication. MONAI Label maintainers shipped a fix on November 18.
What happened
Hackers used Cairn, an AI tool that runs several software agents at once, to break into a server. They told the agents to "obtain shell access," which means gaining control of the machine. The agents found a previously unknown flaw, called a zero-day, in a medical imaging app and used it.
Why it matters to you
AI agents can now find and exploit security holes with little human direction. Amazon warns they can move faster than security teams can respond.
The takeaway
Organizations that run open-source tools should install security updates quickly and use more automated defenses.
What it means for you
Patch MONAI Label to the November 18 release and review exposure of any Cairn-style agent tooling, since the attackers' own open instance exposed them.
Educational institutions running open-source medical imaging software face targeting, and Amazon urges more automated defenses over manual investigation.
The report shows AI agents being used in real intrusions, as reported by Amazon Threat Intelligence, Gambit Security and Google.
Updates
Added source: SecurityWeek.
Added source: The Hacker News.
Label changed from Reported to Confirmed after The Record reported independently.
Added source: The Record.
Newest first. Labels change only by the rules in our standards.