| Framework Home | Research Index | Cross-Pillar Governance | AISM | NEXUS | Dashboard |
Current framework context: AI SAFE² v3.1. This research note preserves its original publication date, evidence, and historical framework references. Use current v3.1 normative control and profile documents for implementation or conformance decisions.
ID: RN-2025-003 | Related Control: [P3.T1.1_ADV], [P4.T2.1_ADV] | Status: Verified
Multi-Agent Swarms (e.g., CrewAI, AutoGen) rely on consensus loops to make decisions.
We treat Swarms as “Distributed Systems” requiring fault tolerance, not just chat interfaces.
A hardware or software “Global Stop” signal that severs network connections for all agents in a swarm simultaneously. This prevents a runaway agent from forking new instances.
Real-time telemetry tracking the “Time-to-Consensus.”
Agents assign reputation scores to peers. If one agent consistently outputs outliers (potential hallucination or compromise), it is mathematically quarantined from the voting pool.
| Previous research note | Research Index | Next research note |
| Framework Home | Cross-Pillar Governance | NEXUS | Challenge Lab |
AI SAFE² v3.1 | Cyber Strategy Institute