CyberLLM: A Multi-Agent LLM Framework for Autonomous Detection and Guarded Response in Automotive Cybersecurity

Software-Defined Vehicles SDVs expand the automotive attack surface across source code, runtime logs, and deployment topologies, while safety constraints forbid autonomous agents from acting without oversight. This paper presents CyberLLM, a multi-agent, LLM-orchestrated framework that autonomous...