Sunday, 21 September, 2025
GPT-4-Powered “MalTerminal” Malware Discovered – Generates Ransomware & Reverse Shells
Cybersecurity researchers from SentinelOne have uncovered “MalTerminal,” the earliest known malware to embed a Large Language Model (GPT-4) for dynamically generating ransomware code or executing a reverse shell. MalTerminal includes an OpenAI API endpoint and Python scripts offering choices between “ransomware” and “reverse shell.” Though no deployments have been observed, it marks a new era in adversary tactics using LLM-embedded malware.
Read full story at The Hacker News