Attack  ·  Glossary

Agentic ransomware

Malware that uses an AI agent to carry out a ransomware attack autonomously — finding its way in, moving through the network, stealing credentials, encrypting data, and wiping backups — all without a human attacker directing each step. The AI replaces the skilled operator, compressing an attack that used to take days into minutes. The first confirmed real-world case (JadePuffer) exploited a flaw in the Langflow AI-orchestration platform.
This marks a qualitative shift in cyber risk: the human expertise required to execute a sophisticated breach is now delegated to software, meaning organisations face a vastly larger pool of potential attackers capable of launching fully automated, enterprise-grade attacks at scale.
Sysdig: JadePuffer — Agentic Ransomware for Automated Database Extortion
Track this in the live feed See how this plays out in real AI security and governance developments.
Open the feed →