Multi-agent systems are software designs where several autonomous agents each handle part of a task, then negotiate, coordinate, and adapt to local conditions. Instead of one central controller making every decision, each agent can monitor its own data, propose actions, and exchange messages with other agents to reach a shared goal.
In cyber security, MAS matters because it can improve resilience and speed, but it also creates new attack paths. In manufacturing or other distributed environments, agents may adjust schedules, routes, or thresholds based on sensor input. If an attacker spoofs telemetry, tampers with messages, or gains control of one over-privileged agent, the system can make bad decisions at scale. Defenders use access controls, message authentication, policy checks, and audit logs to keep agent behavior bounded and explainable. The key security goal is to let agents act locally without letting a compromised agent override the whole system.