[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-84273-en":3,"doc-seo-84273-105":29,"detail-sidebar-cat-0-en-105":90},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},84273,1374391974564,"Clementine","https://ap-avatar.wpscdn.com/avatar/14000253aa45c000a9e?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779874745381141002",8,"Research & Report","Adversarial Social Epistemology for Assemblies of Humans and Large Language Models","Adversarial Social Epistemology (ASE) explains how densely interactive, scaffolded public assertions become targets for strategic distortion. It argues that familiar accounts of bubbles, echo chambers, and misinformation diffusion miss how commitments and entitlements that normally ground trust are exploited. Using epistemic networks with interactive belief structures, ASE specifies trust as confidence in redeemability of commitments. The framework extends to human-machine systems, interpreting LLM failures such as hallucination, sycophancy, evasion, and over-refusal as breakdowns under evaluator pressure.","Adversarial Social Epistemology for Assemblies of Humans and Large Language Models  \nMihnea Moldoveanu and Joel Baum University of Toronto  \nExtended Abstract  \nWe outline an adversarial social epistemology (ASE) for densely interactive communicative landscapes in which public assertions are scaffolded by chains of testimony, inference, institutional certification, and tacit trust. In such landscapes, agents have incentives and affordances to distort, color, omit, fabricate, or strategically under-specify information for private, reputational, rhetorical, or material gains. We argue that these phenomena are not adequately captured by familiar descriptions of epistemic bubbles, echo chambers, or misinformation diffusion. What requires explanation is how communicative agents exploit the commitments and entitlements that normally make scaffolded assertions trustworthy.  \nWe use epistemic networks, or epinets, enriched with interactive belief structures and an inferentialist semantics of assertion, to unpack the trust acts that misinformation breaks. Assertions create upstream commitments to the chains of testimony and inference on which they depend, and downstream commitments to the implications they license. Trust, on this view, is not merely confidence in a speaker’s reliability, but confidence that the speaker could redeem the relevant commitments were entitled interlocutors to ask the questions the assertion makes available.  \nThis apparatus allows us to specify mechanisms by which public or quasi-public communication becomes adversarial. In triadic settings, where a sender addresses a recipient in the presence of observers, communicators can exploit differences in observability, intelligibility, attention, and interactive belief. Poses, demagogical triggers, social-proof covers, smokescreens, plausible ambiguations, decoys, flares, rabbit holes, and haystacks are not merely rhetorical vices. They are maneuvers that use audiences to block, defer, or distort the redemption of inferential commitments.  \nWe show that the same apparatus applies to mixed human-machine networks involving Large Language Models and Large Language Agents. LLM hallucinations, sycophancy, evasive fluency, over-refusal, and dissimulative helpfulness can be understood as machine-regime variants of failures to track and redeem communicative commitments under evaluator pressure. We conclude by sketching three LLM-based epistemic machines: a Brandom Machine for tracking commitments and entitlements, a Hintikka Machine for generating interrogative paths, and a Ramsey Machine for detecting non-veritistic payoffs. Together, they illustrate how adversarial epistemology can be operationalized in environments where trust is necessary, brittle, and easily simulated.  \n1. Introduction  \nEpistemologists study how we should go about figuring out what is true and social epistemologists study social mechanisms underpinning, assisting and sometimes undermining the ‘production function’ of knowledge. But an emphasis on the ways in which networked, scaffolded communications can curtail epistemic agency through the design of environments engineered to exploit the shortcomings of human perception and cognition is of relatively recent vintage [Nguyen, 2023]. The gamification of communication (via ‘like’ and ‘follow’ commands) creates temptations and affordances for people to violate and mis-place their trust and plausibly facilitates the proliferation of epistemic bubbles and echo chambers that undermine the confidence one places in the effects of public dialogue. However, approaches that focus on ‘mean-field’interactions between an epistemic agent (usually a person) and a ‘network’ hosted by an engineered platform necessarily under-weight the range of epistemic mechanisms, interaction types and ‘epistemic games’ agents can play. We (almost) never interact with‘a social field’, even though the observability of communicative acts to (some part of) that field may be salient to ","cbCaisrPUMc7Y008","https://ap.wps.com/l/cbCaisrPUMc7Y008","pdf",421432,1,50,"English","en",105,"# Extended Abstract\n## Introduction","[{\"question\":\"What problem does Adversarial Social Epistemology (ASE) aim to explain?\",\"answer\":\"ASE explains how public or quasi-public assertions become untrustworthy when adversarial agents exploit the commitments and entitlements that usually make such claims reliable.\"},{\"question\":\"How does ASE model trust in scaffolded assertions?\",\"answer\":\"Trust is treated as confidence that the relevant commitments can be redeemed by entitled interlocutors if they ask the questions the assertion makes available.\"},{\"question\":\"How does ASE connect its framework to Large Language Models (LLMs)?\",\"answer\":\"ASE applies the same mechanisms to mixed human-machine networks, interpreting LLM behaviors like hallucination, sycophancy, evasive fluency, over-refusal, and dissimulative helpfulness as failures to track and redeem commitments under evaluator pressure.\"}]",1784194517,126,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":85,"head_meta":87,"extra_data":89,"updated_unix":27},"adversarial-social-epistemology-for-assemblies-of-humans-and-large-language-models","",{"@graph":35,"@context":84},[36,53,67],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/adversarial-social-epistemology-for-assemblies-of-humans-and-large-language-models/84273/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":61,"encodingFormat":60,"isAccessibleForFree":62,"interactionStatistic":63},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-16",true,{"@type":64,"interactionType":65,"userInteractionCount":4},"InteractionCounter",{"@type":66},"ViewAction",{"@type":68,"mainEntity":69},"FAQPage",[70,76,80],{"name":71,"@type":72,"acceptedAnswer":73},"What problem does Adversarial Social Epistemology (ASE) aim to explain?","Question",{"text":74,"@type":75},"ASE explains how public or quasi-public assertions become untrustworthy when adversarial agents exploit the commitments and entitlements that usually make such claims reliable.","Answer",{"name":77,"@type":72,"acceptedAnswer":78},"How does ASE model trust in scaffolded assertions?",{"text":79,"@type":75},"Trust is treated as confidence that the relevant commitments can be redeemed by entitled interlocutors if they ask the questions the assertion makes available.",{"name":81,"@type":72,"acceptedAnswer":82},"How does ASE connect its framework to Large Language Models (LLMs)?",{"text":83,"@type":75},"ASE applies the same mechanisms to mixed human-machine networks, interpreting LLM behaviors like hallucination, sycophancy, evasive fluency, over-refusal, and dissimulative helpfulness as failures to track and redeem commitments under evaluator pressure.","https://schema.org",{"og:url":51,"og:type":86,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":88,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":91},[92,96,100,104,109,113,118,121,126,129,133],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":93,"show_sort_weight":94,"slug":95},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":97,"show_sort_weight":98,"slug":99},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":101,"show_sort_weight":102,"slug":103},"Exam",70,"exam",{"id":105,"doc_module":4,"doc_module_name":45,"category_name":106,"show_sort_weight":107,"slug":108},5,"Comic",60,"comic",{"id":110,"doc_module":4,"doc_module_name":45,"category_name":111,"show_sort_weight":21,"slug":112},6,"Technology","technology",{"id":114,"doc_module":4,"doc_module_name":45,"category_name":115,"show_sort_weight":116,"slug":117},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":119,"slug":120},30,"research-report",{"id":122,"doc_module":4,"doc_module_name":45,"category_name":123,"show_sort_weight":124,"slug":125},9,"Religion & Spirituality",20,"religion-spirituality",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":127,"show_sort_weight":124,"slug":128},"World Cup","world-cup",{"id":130,"doc_module":4,"doc_module_name":45,"category_name":131,"show_sort_weight":130,"slug":132},10,"Lifestyle","lifestyle",{"id":134,"doc_module":4,"doc_module_name":45,"category_name":135,"show_sort_weight":105,"slug":136},19,"General","general"]