Sysdig says the JadePuffer group used an autonomous AI agent to plan and execute an end-to-end ransomware campaign, adapting in real time, stealing credentials and encrypting data while exposing flaws in its financial understanding

A ransomware strain that reportedly lets an AI agent run the attack has arrived. The temptation is to pin it on a famous frontier model. The evidence refuses to let you.

Security firm Sysdig describes an extortion attack where a language model broke in on its own, stole credentials, and destroyed databases. No human appeared to be at the controls.