Red-teaming a network of agents: Understanding what breaks when AI agents interact at scale - Microsoft Research

Safe agents don’t guarantee a safe ecosystem of interconnected agents. Microsoft Research examines what breaks when AI agents interact and why network-level risks require new approaches. Learn more:

Red-teaming a network of agents: Understanding what breaks when AI agents interact at scale - Microsoft Research Skip to main content Research Publications Code & data People Microsoft Research blog Artificial intelligence Audio & acoustics Computer vision Graphics & multimedia Human-computer interaction Human language technologies Search & information retrieval Data platforms and analytics Hardware & devices Programming languages & software engineering Quantum computing Security, privacy & cryptography Systems & networking Algorithms Mathematics Ecology & environment Economics Medical, health

saved by

• Kevin Zhang

related reading

• [2502.15657] Superintelligent Agents Pose Catastrophic Risks: Can Scientist AI Offer a Safer Path?arxiv.org
• Lessons from Moltbook and OpenClaw: The Agentic Internet’s Trust Problem - Irregularirregular.com
• AI 2027ai-2027.com
• Emergent Cyber Behavior: When AI Agents Become Offensive Threat Actors - Irregularirregular.com
• The Rogue Replication Threat Model - METRmetr.org
• [2501.10114] Infrastructure for AI Agentsarxiv.org
• A basic systems architecture for AI agents that do autonomous research — LessWronglesswrong.com
• Welcome to Moltbook - by Zvi Mowshowitzthezvi.substack.com
• Why multi-agent safety is important — LessWronglesswrong.com
• [2411.00114] Project Sid: Many-agent simulations toward AI civilizationarxiv.org
• [2502.14143] Multi-Agent Risks from Advanced AIarxiv.org
• ROGUE:arxiv.org