arXiv:2611.21804v1 Announce Type: new Abstract: AI systems are increasingly deployed in high-stakes decision-making contexts, where interactions among components can introduce fairness risks that remain undetected when auditing components or collective outcomes in isolation. As LLM-based multi-agent systems become more prevalent, understanding how biased information propagates through agent interactions is essential for assessing their fairness. We introduce a controlled bias injection framework that replaces selected agent responses with predefined biased ones, enabling an intervention-based analysis of how network topology and injection configurations shape system behavior. We propose decision divergence metrics that quantify the effects of these interventions at the agent, network, and system levels, and evaluate five topologies on three binary classification tasks using LLMs from the Qwen and Gemma families. Sequential networks show the greatest divergence, whereas fully connected and mesh networks show less. Injection position also matters: in star networks, injecting the central agent has larger effects than injecting peripheral agents. Models further differ in whether divergence follows the injected bias. These findings highlight the role of network structure in the propagation of biased information and motivate topology-aware auditing of fairness risks in multi-agent systems.
Corrupting Ties: Investigating Robustness to Bias Injection Across Multi-Agent Network Topologies
A controlled bias-injection study of LLM-based multi-agent systems found that sequential network topologies show the greatest decision divergence, while fully connected and mesh networks show less, according to arXiv paper 2611.21804v1. The researchers replaced selected agent responses with predefined biased ones and measured effects at agent, network and system levels across five topologies on three binary classification tasks using Qwen and Gemma models. In star networks, injecting the central agent produced larger effects than injecting peripheral agents, and models differed in whether divergence followed the injected bias, findings the authors say motivate topology-aware auditing of fairness risks in multi-agent systems.
Audit fairness at the network level, not per agent: inject biased responses into central and sequential nodes, since those positions produced the largest divergence in this study.
more in playbook: Evals and testing for LLM appsRun your AI side-project on zahid.host
EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.