The "AI-Assisted Pentesting" Suite: Offensive Security has integrated a local, privacy-focused LLM (Large Language Model) helper directly into the terminal. It helps researchers deconstruct complex obfuscated code and suggests exploit payloads without sending data to the cloud.