The Model Knows Better: Defending Medical Multi-Agent Systems Against Prompt Injection
Medical multi-agent systems, in which LLM agents discuss a case, are increasingly studied for clinical decision support.
ProofPaper ↗
Key points
- Prompt injection, text that an attacker adds to a model's input, can manipulate LLMs into wrong decisions.
- How it affects these systems, and how to protect them, remains unclear.
- Existing defenses help only partly: they miss injections that look like ordinary clinical text.
- We therefore propose a training-free defense that verifies the clinical judgments behind the system's decision.
Sources (1)
- [1]The Model Knows Better: Defending Medical Multi-Agent Systems Against Prompt InjectionarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 15, 07:43 PM
Medical multi-agent systems, in which LLM agents discuss a case, are increasingly studied for clinical decision support.
Prompt injection, text that an attacker adds to a model's input, can manipulate LLMs into wrong decisions.
Extractive summary: sentences quoted from the sources.
Before this
- Oct 15, 2026Decomposing Brain-LLM Alignment with Interpretable Semantic Representations
- Oct 15, 2026Incentivizing Tool-Use Reinforcement Learning for Post-Arrival Emergency Care Prediction
- Oct 15, 2026GlobeBench: Evaluating Text-Based World Model Capabilities for Scalable Environment Simulations
- Oct 15, 2026Librarian Agents: Training LLMs to Curate Long-Context Workspaces that Anticipate Future Queries
- Oct 15, 2026JevAgentHarness: Benchmarking, Deploying and Training Decision Models That Share Control with LLMs
- Oct 14, 2026INFUSE: A Unified Surface Benchmark for Indirect Prompt Injection in Agentic Systems