Autonomous Adversary: Red-Teaming in the Age of LLM

Language Model Agents LMAs are emerging as a powerful primitive for augmenting red-team operations. They can support attack planning, adversary emulation, and the orchestration of multi-step activity such as lateral movement, a core enabling capability of advanced persistent threat APT campaigns...