A new type of ransomware, called JADEPUFFER, has emerged that uses an AI model to carry out attacks without human intervention. This agentic threat actor exploited a known vulnerability in a widely used tool and stole credentials, eventually destroying databases. The attack was notable for its speed and ability to correct itself, with the AI model diagnosing and fixing errors in a matter of seconds. Why it matters: this development lowers the barrier to entry for ransomware attacks and highlights the importance of keeping software up to date and addressing old security vulnerabilities.