AI-Powered Ransomware Is the Emerging Threat That Could Bring Down Your Organization

An NYU Tandon research team created Ransomware 3.0 (also called PromptLock), a proof-of-concept AI-powered ransomware that demonstrates how large language models can autonomously orchestrate complete ransomware campaigns. Unlike traditional ransomware with pre-written code, Ransomware 3.0 uses natural language prompts to dynamically generate malicious scripts, making each attack unique and difficult to detect. In a paper with doctoral candidate Md Raz (ECE) as lead author, the team showed the malware analyzes file systems, determines which data to target, and crafts personalized ransom notes by connecting to freely available language models through APIs, revolutionizing how cybercriminals could develop attacks.