Research2026-04-28
Beyond Single-Agent Alignment: Preventing Context-Fragmented Violations in Multi-Agent Systems
Source: Arxiv CS.AI
arXiv:2604.22879v1 Announce Type: cross Abstract: We identify and formalize a novel security risk: Context-Fragmented Violations (CFVs) - a class of policy breaches where individual agent actions appear locally safe and reasonable, yet collectively violate organizational policies because critical...
arxivpapersagentsrag