MADBench: Benchmarking the Security of Multi-Agent Debate

arXiv:2609.39146 · cs.AI · Submitted 2026-09-30 · Read on arXiv

cs.AI

Submitted: 2026-09-30

Updated: 2026-09-30

Terminology

• Benchmarking — 1× in this paper · explained in 4 episodes

Sources

• MAD-Spear: A Conformity-Driven Prompt Injection Attack on Multi-Agent Debate Systems
• Agent Smith: A Single Image Can Jailbreak One Million Multimodal LLM Agents Exponentially Fast
• On the Resilience of LLM-Based Multi-Agent Collaboration with Faulty Agents
• Should we be going MAD? A Look at Multi-Agent Debate Strategies for LLMs
• Multi-Agent Systems Execute Arbitrary Malicious Code
• AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversation
• Misleading the Planner through Deceptive Resumes: Registration-Time Injection in Centralized Multi-Agent Systems
• Universal and Transferable Adversarial Attacks on Aligned Language Models

Related papers

• MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
• Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
• Agents' Last Exam
• The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
• MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
• Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems