A 2026 survey that organizes AI agent systems into a unified taxonomy across deliberation and reasoning, planning and control, and tool calling and environment interaction. Covers agent components (LLM core, memory, world models, planners, tool routers, critics) and single- vs multi-agent orchestration patterns.
Tags
agentic AIfoundationsarchitectureevaluation
At a glance
Published
2026
Jurisdiction
Global
Category
Foundations
Access
Public access
Build your AI governance program
VerifyWise helps you implement AI governance frameworks, track compliance, and manage risk across your AI systems.