Auditing Cascading Risks in Multi-Agent Systems via Semantic-Geometric Co-evolution

TL;DR

This paper introduces a geometric-semantic framework using Ollivier--Ricci Curvature to detect early signs of cascading risks in multi-agent systems, enabling proactive intervention and interpretability.

Contribution

It proposes a novel co-evolutionary analysis method combining semantic signals with graph geometry to identify precursors of cascading failures in MAS.

Findings

Curvature anomalies precede semantic violations by several interaction turns.

The framework enables early detection of cascading risks before explicit semantic issues occur.

Ricci curvature provides interpretable root-cause attribution for system failures.

Abstract

Large Language model (LLM)-based Multi-Agent Systems (MAS) are prone to cascading risks, where early-stage interactions remain semantically fluent and policy-compliant, yet the underlying interaction dynamics begin to distort in ways that amplify latent instability or misalignment. Traditional auditing methods that focus on per-message semantic content are inherently reactive and lagging, failing to capture these early structural precursors. In this paper, we propose a principled framework for cascading-risk detection grounded in semantic--geometric co-evolution. We model MAS interactions as dynamic graphs and introduce Ollivier--Ricci Curvature (ORC) -- a discrete geometric measure -- to characterize information redundancy and bottleneck formation in communication topologies. By coupling semantic flow signals with graph geometry, the framework learns the normal co-evolutionary dynamics of trusted collaboration and treats deviations from this coupled manifold as early-warning signals. Experiments on a suite of cascading-risk scenarios aligned with the risk category demonstrate that curvature anomalies systematically precede explicit semantic violations by several interaction turns, enabling proactive intervention. Furthermore, the local nature of Ricci curvature provides principled interpretability for root-cause attribution, identifying specific agents or links that precipitate the collapse of trustworthy collaboration.