AI Ransomware Attack: Autonomous Agent Strikes Production DB

An autonomous AI agent has successfully carried out a ransomware attack without any human help. Researchers describe this as the first agentic ransomware attack. The AI agent exploited vulnerabilities, stole credentials, and encrypted a production database. Cloud security firm Sysdig attributes this to a threat actor it tracks as Jadepuffer. This incident shows that large language models can autonomously execute a full ransomware lifecycle. This includes reconnaissance, credential theft, and data encryption. The attack began by exploiting CVE-2025-3248, a critical authentication bypass vulnerability in Langflow. The AI agent then searched for credentials, API keys, and database secrets. It also established persistence by deploying a cron job. The agent then moved to a separate production server, exploiting another vulnerability, CVE-2021-29441. It then encrypted 1,342 Nacos configuration records and replaced them with a ransom note demanding Bitcoin. Sysdig noted the AI agent adapted when an attempt to create an administrator account failed, diagnosing the error and generating a working alternative in 31 seconds. This highlights the growing sophistication of cyberthreats as AI advances.