Multi-agent system
Also known as: MAS, agent network, multi-agent architecture
A production deployment in which two or more autonomous AI agents share state, share context, or coordinate on tasks. The threshold for governance purposes is the inter-agent communication path: if information flows from one agent's output to another agent's context window, memory, or tool inputs, the deployment is multi-agent.
Most enterprise multi-agent deployments resolve to one of three orchestration patterns: hierarchical (one orchestrator delegates), peer-to-peer (agents talk directly), or broker-mediated (a central policy enforces). Broker-mediated is the right default for high-risk deployments because every cross-agent message is logged and policy-checked at one chokepoint. Peer-to-peer is research-grade — production above the low-risk threshold should not adopt it without strengthening the audit substrate beyond vendor-native baseline.