Trust but Verify: Mitigating Medical Hallucinations via Post-Hoc Adversarial Auditing and Multi-Agent Feedback Loops

• URL: http://arxiv.org/abs/2606.14149v1
• Date: Fri, 12 Jun 2026 06:21:19 GMT
• Title: Trust but Verify: Mitigating Medical Hallucinations via Post-Hoc Adversarial Auditing and Multi-Agent Feedback Loops
• Abstract summary: Large Language Models (LLMs) are increasingly deployed in healthcare settings, yet their tendency to hallucinate poses risks when clinical decisions are involved.<n>This study examine whether LLMs recommend recently banned or withdrawn pharmaceuticals when answering clinical questions.<n>We developed a five-agent "Trust but verify" system using a single LLM backbone.
• Score: 0.07249400282852114
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large Language Models (LLMs) are increasingly deployed in healthcare settings, yet their tendency to hallucinate poses risks when clinical decisions are involved. This study examine whether LLMs recommend recently banned or withdrawn pharmaceuticals when answering clinical questions and tests an agent-based method for reducing such errors. We developed a five-agent "Trust but Verify" system using a single LLM backbone. To measure regulatory knowledge obsolescence, we created an adversarial dataset of 103 clinical MCQs where historically correct answers now refer to banned substances. This scale ensures statistical significance across various therapeutic classes. We evaluated three open-access model families (GPT-OSS, Llama-3, Falcon-3) under vanilla and agentic conditions. Performance was measured via pointwise score, label accuracy, Hallucination Error Rate (HER), and Component Fidelity (CF) score. We also observed clinical safety regression in proprietary models. In default configurations, all models showed high hallucination rates, consistently selecting banned drugs that matched training data patterns. Our proposed agentic architecture reduced HER by approximately 53% across models. Pointwise scores shifted from -0.25 (unsafe recommendation) toward 0.0 (appropriate refusal). The safety audit intercepted dangerous outputs even when models' parametric knowledge favored the banned substance. The proposed multi-agent framework offers a model-agnostic method for enforcing regulatory compliance that prioritizes patient safety over fluent text generation. Our work demonstrates a practical approach for deploying autonomous AI systems in safety-critical healthcare settings. It shows how real-time regulatory data can be integrated into LLM pipelines to support clinical decision-making.

Related papers

• The Mirage of LLM Guardrails: A Case Study in AI-Assisted Medical Note Manipulation [7.091077137729042]
  We use AI-assisted medical note manipulation as a concrete case study.<n>First, we develop a reproducible manipulation pipeline that takes publicly available seed medical note templates.<n>Second, we conduct a systematic empirical evaluation of LLM guardrail robustness for medical note manipulation.<n>Third, we utilize a combination of automated metrics and human annotation-based metrics to assess the correctness of requested manipulations.
  arXiv  Detail & Related papers  (2026-07-26T07:46:28Z)
• MedGuards: Multi-Agent System for Reliable Medical Error Detection and Correction [7.8939854491604455]
  MedGuards is a new framework that treats medical error detection and correction as a multi-agent in-context learning task.<n> Specialized agents separately detect, localize, and correct errors, while a confidence-guided arbitration mechanism resolves disagreements.<n>Our aim is to enable safer deployment of Large Language Models in real-world healthcare applications.
  arXiv  Detail & Related papers  (2026-06-24T10:07:59Z)
• Beyond Scalar Scores: Exploring LLM-based Metrics for Clinical Significance Evaluation in Radiology Reports [49.5225801722164]
  Existing metrics obscure this requirement by reducing report quality to a medically ungrounded scalar.<n>We study this boundary using ReEvalMed benchmark as testbed and evaluate metric-level clinical significance.
  arXiv  Detail & Related papers  (2026-06-17T08:10:30Z)
• Agentic AI-based Framework for Mitigating Premature Diagnostic Handoff and Silent Hallucination in Healthcare Applications [17.69274632062373]
  Open-ended conversational agents remain prone to two critical failure modes: premature diagnostic handoff and silent clinical hallucinations.<n>We propose a multi-agent framework that addresses both issues by replacing LLM-as-a-judge'' routing with deterministic orchestration constraints.<n>We evaluate the system using simulated patient agents powered by the llama-3.1-70b-instruct model on 150 test cases.
  arXiv  Detail & Related papers  (2026-06-16T15:39:19Z)
• CARE: A Conformal Safety Layer for Medical Summarization [24.00849890172944]
  Large language models (LLMs) are increasingly used for medical summarization.<n>Existing error-detection methods produce or uncalibrated scores, providing no formal control over missed errors.<n>We introduce Conformal Assessment for Risk Evaluation (CARE), a model-agnostic safety layer that uses conformal risk control to overlay omission and hallucination flags onto calibrated summaries.
  arXiv  Detail & Related papers  (2026-06-08T03:10:45Z)
• Med-HEAL: Analyzing and Mitigating Hallucinations in Medical LLMs with Hallucination-Aware In-Context Learning [8.322191814123315]
  Hallucinations in medical large language models pose serious risks for clinical decision support.<n>We introduce Med-HEAL, a framework for systematically identifying, analyzing, and mitigating hallucinations in medical LLMs.
  arXiv  Detail & Related papers  (2026-05-31T15:43:42Z)
• ProtoMedAgent: Multimodal Clinical Interpretability via Privacy-Aware Agentic Workflows [7.27267618572654]
  ProtoMedAgent formalizes multimodal clinical reporting as an iterative, zero-gradient test-time optimization problem.<n>Online generation is strictly constrained by exact set-theoretic differentials and a reflective Scribe-Critic loop.<n>ProtoMedAgent achieves 91.2% Comparison Set Faithfulness where it fundamentally outperforms standard RAG (46.2%)
  arXiv  Detail & Related papers  (2026-05-13T20:57:37Z)
• A Federated and Parameter-Efficient Framework for Large Language Model Training in Medicine [59.78991974851707]
  Large language models (LLMs) have demonstrated strong performance on medical benchmarks, including question answering and diagnosis.<n>Most medical LLMs are trained on data from a single institution, which faces limitations in generalizability and safety in heterogeneous systems.<n>We introduce the model-agnostic and parameter-efficient federated learning framework for adapting LLMs to medical applications.
  arXiv  Detail & Related papers  (2026-01-29T18:48:21Z)
• Conformal Lesion Segmentation for 3D Medical Images [82.92159832699583]
  We propose a risk-constrained framework that calibrates data-driven thresholds via conformalization to ensure the test-time FNR remains below a target tolerance.<n>We validate the statistical soundness and predictive performance of CLS on six 3D-LS datasets across five backbone models, and conclude with actionable insights for deploying risk-aware segmentation in clinical practice.
  arXiv  Detail & Related papers  (2025-10-19T08:21:00Z)
• A Granular Study of Safety Pretraining under Model Abliteration [64.24346997570275]
  We study model abliteration, a lightweight projection technique designed to remove refusal-sensitive directions.<n>We issue 100 prompts with balanced harmful and harmless cases, classify responses as **Refusal** or **Non-Refusal** using multiple judges, and validate judge fidelity.<n>Our study produces a checkpoint-level characterization of which data-centric safety components remain robust under abliteration.
  arXiv  Detail & Related papers  (2025-10-03T07:01:45Z)
• From Scores to Steps: Diagnosing and Improving LLM Performance in Evidence-Based Medical Calculations [45.414878840652115]
  Large language models (LLMs) have demonstrated promising performance on medical benchmarks.<n>However, their ability to perform medical calculations remains underexplored and poorly evaluated.<n>In this work, we revisit medical calculation evaluation with a stronger focus on clinical trustworthiness.
  arXiv  Detail & Related papers  (2025-09-20T09:10:26Z)
• Beyond Benchmarks: Dynamic, Automatic And Systematic Red-Teaming Agents For Trustworthy Medical Language Models [87.66870367661342]
  Large language models (LLMs) are used in AI applications in healthcare.<n>Red-teaming framework that continuously stress-test LLMs can reveal significant weaknesses in four safety-critical domains.<n>A suite of adversarial agents is applied to autonomously mutate test cases, identify/evolve unsafe-triggering strategies, and evaluate responses.<n>Our framework delivers an evolvable, scalable, and reliable safeguard for the next generation of medical AI.
  arXiv  Detail & Related papers  (2025-07-30T08:44:22Z)
• Correctness Coverage Evaluation for Medical Multiple-Choice Question Answering Based on the Enhanced Conformal Prediction Framework [2.9599960287815144]
  Large language models (LLMs) are increasingly adopted in medical question-answering (QA) scenarios.<n>LLMs can generate hallucinations and nonfactual information, undermining their trustworthiness in high-stakes medical tasks.<n>This paper proposes an enhanced Conformal Prediction framework for medical multiple-choice question-answering tasks.
  arXiv  Detail & Related papers  (2025-03-07T15:22:10Z)
• SORRY-Bench: Systematically Evaluating Large Language Model Safety Refusal [64.9938658716425]
  SORRY-Bench is a proposed benchmark for evaluating large language models' (LLMs) ability to recognize and reject unsafe user requests.<n>First, existing methods often use coarse-grained taxonomy of unsafe topics, and are over-representing some fine-grained topics.<n>Second, linguistic characteristics and formatting of prompts are often overlooked, like different languages, dialects, and more -- which are only implicitly considered in many evaluations.
  arXiv  Detail & Related papers  (2024-06-20T17:56:07Z)
• Large Language Model Distilling Medication Recommendation Model [58.94186280631342]
  We harness the powerful semantic comprehension and input-agnostic characteristics of Large Language Models (LLMs)<n>Our research aims to transform existing medication recommendation methodologies using LLMs.<n>To mitigate this, we have developed a feature-level knowledge distillation technique, which transfers the LLM's proficiency to a more compact model.
  arXiv  Detail & Related papers  (2024-02-05T08:25:22Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.