AI agent carries out full ransomware attack without human assistance
Security experts at Sysdig say they have recorded the initial instance of agentic ransomware, which is a cyberattack performed from start to finish by an AI agent without any human guidance. The AI, named JADEPUFFER, gained access to a server connected to the internet, breached a database, and stole files for the purpose of extortion. Remarkably, the agent was also able to correct its own coding mistakes within 31 seconds.
by shortkt.com✓