The article explores the concept of developing C2-less malware using large language models (LLMs) for autonomous decision-making and exploitation. It discusses the implications of such technology, particularly through a malware example called "PromptLock," which utilizes LLMs to generate and execute code without human intervention. The author proposes a proof of concept for self-contained malware capable of exploiting misconfigured services on a target system.