#multi-agent-systems

The Emergent Risks of Autonomous Agent Swarms

Autonomous LLM agents are forming complex systems where safety risks emerge at the collective level, rendering individual-agent security policies ineffective.