TRiSM for Agentic AI: A Review of Trust, Risk, and Security Management in LLM-based Agentic Multi-Agent Systems

Synopsis

Extended TRiSM framework — explainability, ModelOps, security, privacy, lifecycle governance — for LLM-based multi-agent systems, with a multi-agent risk taxonomy and two novel metrics (Component Synergy Score, Tool Utilization Efficacy).

Keywords

·TRiSM ·multi-agent systems ·governance ·risk taxonomy ·lifecycle

Related entries

• A Co-Evolutionary Theory of Human-AI Coexistence: Mutualism, Governance, and Dynamics in Complex Societies
  April 24, 2026 · arXiv
• AGENTSAFE: A Unified Framework for Ethical Assurance and Governance in Agentic AI
  December 2, 2025 · arXiv
• Agentic AI has a Human Oversight Problem
  September 15, 2025 · SSRN preprint 2025
• The AI Agent Index
  February 3, 2025 · arXiv
• On the Quest for Effectiveness in Human Oversight: Interdisciplinary Perspectives
  April 5, 2024 · FAccT 2024
• Secret Collusion among AI Agents: Multi-Agent Deception via Steganography
  February 12, 2024 · NeurIPS 2024