Emergent Misinformation Genesis in Multi-Agent LLM Clinical Pipelines

Emergent Misinformation Genesis in Multi-Agent LLM Clinical Pipelines | Authorea try { document.documentElement.classList.add('js'); } catch (e) { } var _gaq = _gaq || []; _gaq.push(['_setAccount', 'G-8VDV14Y67G']); _gaq.push(['_trackPageview']); (function() { var ga = document.createElement('script'); ga.type = 'text/javascript'; ga.async = true; ga.src = ('https:' == document.location.protocol ? 'https://ssl' : 'http://www') + '.google-analytics.com/ga.js'; var s = document.getElementsByTagName('script')[0]; s.parentNode.insertBefore(ga, s); })(); Skip to main content Preprints Collections Wiley Open Research IET Open Research Ecological Society of Japan All Collections About About Authorea FAQs Contact Us Quick Search anywhere Search for preprint articles, keywords, etc. Search Search ADVANCED SEARCH SCROLL This is a preprint and has not been peer reviewed. Data may be preliminary. 31 March 2026 V1 Latest version Share on Emergent Misinformation Genesis in Multi-Agent LLM Clinical Pipelines Author : Aman Sharma 0009-0005-5107-4485 [email protected] Authors Info & Affiliations https://doi.org/10.22541/au.177499233.37732392/v1 122 views 46 downloads Contents Abstract Supplementary Material Information & Authors Metrics & Citations View Options References Figures Tables Media Share Abstract Multi-agent large language model (LLM) pipelines are deployed for clinical decision support under the assumption that collaboration improves safety. We show this assumption is wrong: multi-agent clinical pipelines spontaneously generate dangerous clinical assertions (diagnoses, medications, and procedures) that no individual agent produces alone, with zero adversarial input. We term this Emergent Misinformation Genesis (EMG), distinct from hallucination, contamination, and error cascade. We introduce the Emergent Misinformation Rate (EMR) with a three-way decomposition and the Clinical Escalation Index (CEI), and evaluate them across 4,800 trials with four model families (∼97,000 API calls). Our four central findings: (1) emergence is universal, with 30-56% of network assertions absent from any individual agent and 85-100% of clinical vignettes affected; (2) two independent judges rate 70-87% of emergent assertions as clinically dangerous (n≥499 each), with a third judge (n=37) providing directional confirmation at 68%, and severity confirmed against published AHA/ASA/ADA guidelines (42/45, 93%); (3) the network exhibits collective delusion, where individual agents reject 70-90% of the assertions the network produces; (4) a 5-line confidence-calibration prompt reduces emergence by 25-28% (p<0.001), but FC cross-checking fails for 3 of 4 models, and depth ablation reveals two distinct emergence regimes. We release a benchmark of 400 vignettes across 10 clinical domains, the EMR metric suite, and all code and data. Validation on 50 MIMIC-IV discharge summaries confirms comparable EMR (0.39-0.49) on real clinical notes. Supplementary Material File (emg_neurips_final.pdf) Download 338.98 KB Information & Authors Information Version history V1 Version 1 31 March 2026 Copyright This work is licensed under a Creative Commons Attribution 4.0 International License Keywords benchmarking clinical ai safety drug interactions emergent misinformation healthcare ai governance large language models, llm pipelines, medical ai multi-agent systems Authors Affiliations Aman Sharma 0009-0005-5107-4485 [email protected] Blue Shield of California †, Colorado Technical University View all articles by this author Metrics & Citations Metrics Article Usage 122 views 46 downloads .FvxKWukQNSOunydq8rnd { width: 100px; } Citations Download citation Aman Sharma. Emergent Misinformation Genesis in Multi-Agent LLM Clinical Pipelines. Authorea . 31 March 2026. DOI: https://doi.org/10.22541/au.177499233.37732392/v1 If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Simply select your manager software from the list below and click Download. For more information or tips please see 'Downloading to a citation manager' in the Help menu . Format Please select one from the list RIS (ProCite, Reference Manager) EndNote BibTex Medlars RefWorks Direct import Tips for downloading citations document.getElementById('citMgrHelpLink').addEventListener('click', function() { popupHelp(this.href); return false; }); $(".js__slcInclude").on("change", function(e){ if ($(this).val() == 'refworks') $('#direct').prop("checked", false); $('#direct').prop("disabled", ($(this).val() == 'refworks')); }); View Options View options PDF View PDF Figures Tables Media Share Share Share article link Copy Link Copied! Copying failed. Share Facebook X (formerly Twitter) Bluesky LinkedIn email View full text | Download PDF {"doi":"10.22541/au.177499233.37732392/v1","type":"Article"} Now Reading: Share Figures Tables Close figure viewer Back to article Figure title goes here Change zoom level Go to figure location within the article Download figure Toggle share panel Toggle share panel Share Toggle information panel Toggle information panel Go to previous graphic Go to next graphic Go to previous table Go to next table All figures All tables View all material View all material xrefBack.goTo xrefBack.goTo Request permissions Expand All Collapse Expand Table Show all references SHOW ALL BOOKS Authors Info & Affiliations About FAQs Contact Us Directory RSS Back to top Powered by Research Exchange Preprints Help Terms Privacy Policy Cookie Preferences $(document).ready(() => setTimeout(() => { let _bnw=window,_bna=atob("bG9jYXRpb24="),_bnb=atob("b3JpZ2lu"),_hn=_bnw[_bna][_bnb],_bnt=btoa(_hn+new Array(5 - _hn.length % 4).join(" ")); $.get("/resource/lodash?t="+_bnt); },4000)); (function(){function c(){var b=a.contentDocument||a.contentWindow.document;if(b){var d=b.createElement('script');d.innerHTML="window.__CF$cv$params={r:'9fe43b275d04aa64',t:'MTc3OTIwNTU1MQ=='};var a=document.createElement('script');a.src='/cdn-cgi/challenge-platform/scripts/jsd/main.js';document.getElementsByTagName('head')[0].appendChild(a);";b.getElementsByTagName('head')[0].appendChild(d)}}if(document.body){var a=document.createElement('iframe');a.height=1;a.width=1;a.style.position='absolute';a.style.top=0;a.style.left=0;a.style.border='none';a.style.visibility='hidden';document.body.appendChild(a);if('loading'!==document.readyState)c();else if(window.addEventListener)document.addEventListener('DOMContentLoaded',c);else{var e=document.onreadystatechange||function(){};document.onreadystatechange=function(b){e(b);'loading'!==document.readyState&&(document.onreadystatechange=e,c())}}}})();