{"id":6102,"date":"2026-10-19T16:48:50","date_gmt":"2026-10-19T16:48:50","guid":{"rendered":"https:\/\/i10xnews.kinsta.cloud\/?p=6102"},"modified":"2026-10-19T17:15:00","modified_gmt":"2026-10-19T17:15:00","slug":"memadapter-memory-induced-sycophancy-agents","status":"publish","type":"post","link":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents","title":{"rendered":"Correct Memories Still Make AI Agents Agree With You Too Much"},"content":{"rendered":"\n<h2 class=\"wp-block-heading\">\u26a1 Quick Take<\/h2>\n\n<ul class=\"wp-block-list\"><li><strong>The finding:<\/strong> When agents retrieve only factually correct, task-relevant memories, overall accuracy still falls from 96.0% to 77.3%. Among answers that were correct without memory, 19.4% flip to wrong once memory is added \u2014 always in the direction of the retrieved memory.<\/li><li><strong>The fix:<\/strong> MemAdapter uses counterfactual induction, context-aware reflection and evidence-based reasoning to decide how much influence each memory should have.<\/li><li><strong>What it means:<\/strong> Filtering bad memories is not enough. Even good memories can warrant different influence in different contexts.<\/li><\/ul>\n\n\n<p class=\"wp-block-paragraph\">Long-term memory lets agents personalize and work across sessions. It can also make them sycophantic: over-aligning with a user&#8217;s historical beliefs even when those beliefs conflict with current evidence. Most fixes assume the problem is biased or incorrect memories. Researchers from Jilin University and Xiamen University show that objective, correct memories can still induce sycophancy.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1344\" height=\"768\" src=\"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/10\/memadapter-memory-induced-sycophancy-agents-accuracy.webp\" alt=\"Dumbbell chart: accuracy falls from 96 to 77.3 overall when correct memories are added, and similarly on three benchmarks from 96 to 68, 100 to 84 and 92 to 80.\" class=\"wp-image-6101\" srcset=\"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/memadapter-memory-induced-sycophancy-agents-accuracy.webp 1344w, https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/memadapter-memory-induced-sycophancy-agents-accuracy-300x171.webp 300w, https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/memadapter-memory-induced-sycophancy-agents-accuracy-1024x585.webp 1024w, https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/memadapter-memory-induced-sycophancy-agents-accuracy-768x439.webp 768w\" sizes=\"auto, (max-width: 1344px) 100vw, 1344px\" \/><figcaption class=\"wp-element-caption\">i10X chart based on data from Ning et al. (2026), arXiv:2610.05162, Figure 2a preliminary study (CC BY 4.0).<\/figcaption><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83d\udd2c What the study shows<\/h2>\n\n\n<p class=\"wp-block-paragraph\">A preliminary study using only factually correct and task-relevant memories found accuracy drops across three benchmarks (96\u219268%, 100\u219284%, 92\u219280%). The same memory can help in one context and hurt in another. MemAdapter therefore adapts how memories are integrated rather than only filtering what is stored.<\/p>\n\n\n<ul class=\"wp-block-list\"><li><strong>Large gains on hard controls:<\/strong> With Mem0, MemAdapter improved Contextual Scope Control and Memory\u2013Evidence Conflict by 61.67 and 54.00 points; with LightMem, by 57.00 and 92.67 points.<\/li><li><strong>Stabilizes across systems:<\/strong> Raw accuracy across memory backends ranged from 37.03 (LightMem) to 71.71 (A-MEM). With MemAdapter the range narrowed to 85.48\u201390.45.<\/li><li><strong>Beats competing interventions:<\/strong> On A-MEM, MemAdapter reached 91.22% on When to Use Memory and 89.39% on How to Use Memory, ahead of the strongest competing intervention by 0.44 and 2.33 points.<\/li><li><strong>Quality scores rise:<\/strong> Across memory systems, MemAdapter obtained the highest average human-style quality scores, with gains of 0.29 to 0.56 points.<\/li><\/ul>\n\n\n<h2 class=\"wp-block-heading\">\u2705 Why it matters: memory is not neutral context<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Product teams treat memory as a pure upgrade for personalization. This paper shows retrieval can systematically pull answers toward past user views, even when those views are outdated relative to new evidence.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Filtering at write time cannot solve context-dependent risk. The same fact needs different weight tomorrow than today.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83e\uddea Method in brief<\/h2>\n\n\n<p class=\"wp-block-paragraph\">MemAdapter&#8217;s three stages: Counterfactual Induction (probe how a memory could mislead), Context-Aware Reflection (calibrate influence for the current task), Evidence-Based Reasoning (ground the answer while preserving legitimate memory use). Evaluated on three benchmarks across memory systems including A-MEM, Mem0 and LightMem. Code: github.com\/DEEP-JLU\/MemAdapter.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\u2753 What remains open<\/h2>\n\n<ul class=\"wp-block-list\"><li><strong>Benchmarked scenarios:<\/strong> Results are on constructed sycophancy and memory-use benchmarks, not production chat logs.<\/li><li><strong>Extra inference cost:<\/strong> Counterfactual and reflection stages add latency; the paper reports P95\/P99 latency metrics.<\/li><\/ul>\n\n\n<h2 class=\"wp-block-heading\">\ud83e\udded Context<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Related to broader work on sycophancy in LLMs and long-horizon agent memory. See <a href=\"\/news\/category\/studies\/safety-alignment\">Safety &amp; Alignment<\/a> and <a href=\"\/news\/category\/studies\/agent-research\">Agents<\/a>.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83d\udee0\ufe0f What it means for agent memory systems<\/h2>\n\n<ul class=\"wp-block-list\"><li><strong>Do not auto-trust retrieval:<\/strong> Gate memory influence by current evidence, not just relevance.<\/li><li><strong>Test memory-on vs memory-off:<\/strong> Measure flip rates where correct answers become wrong.<\/li><li><strong>Adapt, don&#8217;t only filter:<\/strong> Correct memories still need context-aware weighting.<\/li><\/ul>\n\n\n<h2 class=\"wp-block-heading\">\u270d\ufe0f About the analysis<\/h2>\n\n\n<p class=\"wp-block-paragraph\">This article is based on the full text of the preprint &#8220;Memadapter: Counterfactual Adaptation Against Memory-induced Sycophancy&#8221; (arXiv:2610.05162v1, submitted October 4, 2026). All numbers are taken from the paper&#8217;s text, tables and figures; i10X did not rerun the experiments. The chart was redrawn by i10X from Figure 2a preliminary study of the paper, which is published under a CC BY 4.0 license.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">\ud83d\udd2d i10x Perspective<\/h2>\n\n\n<p class=\"wp-block-paragraph\">Sycophancy usually sounds like flattery. Here it is more structural: yesterday&#8217;s user belief, stored as a &#8220;correct&#8221; memory, quietly overrides today&#8217;s facts. MemAdapter&#8217;s useful idea is to treat memory influence as something to reason about, not a binary include\/exclude. That is closer to how careful people use notes \u2014 and closer to what long-horizon agents will need.<\/p>\n\n","protected":false},"excerpt":{"rendered":"<p>Even factually correct retrieved memories cut accuracy from 96% to 77% by pulling agents toward users&#8217; past views. MemAdapter recovers reliability across three memory systems.<\/p>\n","protected":false},"author":2,"featured_media":6252,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"paper_title":"Memadapter: Counterfactual Adaptation Against Memory-induced Sycophancy","paper_url":"https:\/\/arxiv.org\/abs\/2610.05162","pdf_url":"https:\/\/arxiv.org\/pdf\/2610.05162v1","doi":"","arxiv_id":"2610.05162","authors":"Ruqing Ning, Haibo Meng, Zhishang Xiang, Zerui Chen, Jinsong Su, Xin Wang, Qinggang Zhang","institutions":"Jilin University; Xiamen University","paper_date":"2026-10-04","venue":"arXiv (cs.AI), version 1","peer_reviewed":false,"code_url":"https:\/\/github.com\/DEEP-JLU\/MemAdapter","data_url":"","funding_coi":"No competing-interest statement found in the excerpt.","figure_license":"Chart drawn by i10X from Figure 2a preliminary study (CC BY 4.0)","paper_license":"CC BY 4.0","replication":"no","study_models":"LLM agents with memory backends A-MEM, Mem0, LightMem (and baselines)","study_sample":"Three memory\/sycophancy benchmarks with preliminary accuracy study and full MemAdapter evaluation","footnotes":"","_yoast_wpseo_primary_category":""},"categories":[54,52],"tags":[],"companies":[],"products":[],"class_list":["post-6102","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-safety-alignment","category-studies"],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v28.7 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>MemAdapter: Fighting Memory-Induced Sycophancy in Agents - i10X News<\/title>\n<meta name=\"description\" content=\"Study: correct memories still induce sycophancy, cutting accuracy from 96% to 77%. MemAdapter adapts memory use and narrows the gap.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"MemAdapter: Fighting Memory-Induced Sycophancy in Agents - i10X News\" \/>\n<meta property=\"og:description\" content=\"Study: correct memories still induce sycophancy, cutting accuracy from 96% to 77%. MemAdapter adapts memory use and narrows the gap.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents\" \/>\n<meta property=\"og:site_name\" content=\"i10X News\" \/>\n<meta property=\"article:published_time\" content=\"2026-10-19T16:48:50+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2026-10-19T17:15:00+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/memadapter-memory-induced-sycophancy-agents-title.webp\" \/>\n\t<meta property=\"og:image:width\" content=\"1344\" \/>\n\t<meta property=\"og:image:height\" content=\"768\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/webp\" \/>\n<meta name=\"author\" content=\"Christopher Ort\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@i10x_ai\" \/>\n<meta name=\"twitter:site\" content=\"@i10x_ai\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"Christopher Ort\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"3 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"NewsArticle\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents\"},\"author\":{\"name\":\"Christopher Ort\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/person\\\/d410bb7ee72febefca2c9c369939219a\"},\"headline\":\"Correct Memories Still Make AI Agents Agree With You Too Much\",\"datePublished\":\"2026-10-19T16:48:50+00:00\",\"dateModified\":\"2026-10-19T17:15:00+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents\"},\"wordCount\":629,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#organization\"},\"image\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/memadapter-memory-induced-sycophancy-agents-title.webp\",\"articleSection\":[\"Safety &amp; Alignment\",\"Studies\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents\",\"name\":\"MemAdapter: Fighting Memory-Induced Sycophancy in Agents - i10X News\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/memadapter-memory-induced-sycophancy-agents-title.webp\",\"datePublished\":\"2026-10-19T16:48:50+00:00\",\"dateModified\":\"2026-10-19T17:15:00+00:00\",\"description\":\"Study: correct memories still induce sycophancy, cutting accuracy from 96% to 77%. MemAdapter adapts memory use and narrows the gap.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents#primaryimage\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/memadapter-memory-induced-sycophancy-agents-title.webp\",\"contentUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/memadapter-memory-induced-sycophancy-agents-title.webp\",\"width\":1344,\"height\":768,\"caption\":\"Illustration: i10X\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/memadapter-memory-induced-sycophancy-agents#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/news.i10x.ai\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Correct Memories Still Make AI Agents Agree With You Too Much\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#website\",\"url\":\"https:\\\/\\\/news.i10x.ai\\\/\",\"name\":\"i10X News\",\"description\":\"Stay updated with the latest news and announcements from i10X.\",\"publisher\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/news.i10x.ai\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#organization\",\"name\":\"i10X\",\"url\":\"https:\\\/\\\/news.i10x.ai\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/i10x-logo-180.png\",\"contentUrl\":\"https:\\\/\\\/i10xnews.kinsta.cloud\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/i10x-logo-180.png\",\"width\":180,\"height\":180,\"caption\":\"i10X\"},\"image\":{\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/logo\\\/image\\\/\"},\"sameAs\":[\"https:\\\/\\\/x.com\\\/i10x_ai\",\"https:\\\/\\\/www.linkedin.com\\\/company\\\/i10x-ai\",\"https:\\\/\\\/instagram.com\\\/i10x.ai\",\"https:\\\/\\\/www.youtube.com\\\/@i10X_ai_official\"]},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/news.i10x.ai\\\/#\\\/schema\\\/person\\\/d410bb7ee72febefca2c9c369939219a\",\"name\":\"Christopher Ort\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"contentUrl\":\"https:\\\/\\\/i10x.ai\\\/news\\\/wp-content\\\/uploads\\\/2026\\\/10\\\/christopher-ort-192.webp\",\"caption\":\"Christopher Ort\"},\"description\":\"Christopher Ort handles SEO, AI agent workflows and PR & partnerships at i10x.ai, the all-in-one AI workspace built by i10X Global Pte. Ltd. in Singapore. He is a graduate of Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU). He is the author of \u201cThe Style Penalty\u201d, a study published on SSRN that analyses 1,576 evaluation decisions by four frontier AI models to show how AI-written CVs fare in automated hiring screens; the research has been covered by heise online and t3n.\",\"url\":\"https:\\\/\\\/i10x.ai\\\/news\\\/author\\\/christopher-ort\",\"sameAs\":[\"https:\\\/\\\/i10x.ai\\\/blog\\\/author\\\/christopher-ort\"],\"worksFor\":{\"@id\":\"https:\\\/\\\/i10x.ai\\\/news\\\/#organization\"},\"alumniOf\":{\"@type\":\"EducationalOrganization\",\"name\":\"Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU)\",\"url\":\"https:\\\/\\\/www.fau.eu\"}}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"MemAdapter: Fighting Memory-Induced Sycophancy in Agents - i10X News","description":"Study: correct memories still induce sycophancy, cutting accuracy from 96% to 77%. MemAdapter adapts memory use and narrows the gap.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents","og_locale":"en_US","og_type":"article","og_title":"MemAdapter: Fighting Memory-Induced Sycophancy in Agents - i10X News","og_description":"Study: correct memories still induce sycophancy, cutting accuracy from 96% to 77%. MemAdapter adapts memory use and narrows the gap.","og_url":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents","og_site_name":"i10X News","article_published_time":"2026-10-19T16:48:50+00:00","article_modified_time":"2026-10-19T17:15:00+00:00","og_image":[{"width":1344,"height":768,"url":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/memadapter-memory-induced-sycophancy-agents-title.webp","type":"image\/webp"}],"author":"Christopher Ort","twitter_card":"summary_large_image","twitter_creator":"@i10x_ai","twitter_site":"@i10x_ai","twitter_misc":{"Written by":"Christopher Ort","Est. reading time":"3 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"NewsArticle","@id":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents#article","isPartOf":{"@id":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents"},"author":{"name":"Christopher Ort","@id":"https:\/\/news.i10x.ai\/#\/schema\/person\/d410bb7ee72febefca2c9c369939219a"},"headline":"Correct Memories Still Make AI Agents Agree With You Too Much","datePublished":"2026-10-19T16:48:50+00:00","dateModified":"2026-10-19T17:15:00+00:00","mainEntityOfPage":{"@id":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents"},"wordCount":629,"commentCount":0,"publisher":{"@id":"https:\/\/news.i10x.ai\/#organization"},"image":{"@id":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents#primaryimage"},"thumbnailUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/memadapter-memory-induced-sycophancy-agents-title.webp","articleSection":["Safety &amp; Alignment","Studies"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents#respond"]}]},{"@type":"WebPage","@id":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents","url":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents","name":"MemAdapter: Fighting Memory-Induced Sycophancy in Agents - i10X News","isPartOf":{"@id":"https:\/\/news.i10x.ai\/#website"},"primaryImageOfPage":{"@id":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents#primaryimage"},"image":{"@id":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents#primaryimage"},"thumbnailUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/memadapter-memory-induced-sycophancy-agents-title.webp","datePublished":"2026-10-19T16:48:50+00:00","dateModified":"2026-10-19T17:15:00+00:00","description":"Study: correct memories still induce sycophancy, cutting accuracy from 96% to 77%. MemAdapter adapts memory use and narrows the gap.","breadcrumb":{"@id":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents#primaryimage","url":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/memadapter-memory-induced-sycophancy-agents-title.webp","contentUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/memadapter-memory-induced-sycophancy-agents-title.webp","width":1344,"height":768,"caption":"Illustration: i10X"},{"@type":"BreadcrumbList","@id":"https:\/\/i10x.ai\/news\/memadapter-memory-induced-sycophancy-agents#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/news.i10x.ai\/"},{"@type":"ListItem","position":2,"name":"Correct Memories Still Make AI Agents Agree With You Too Much"}]},{"@type":"WebSite","@id":"https:\/\/news.i10x.ai\/#website","url":"https:\/\/news.i10x.ai\/","name":"i10X News","description":"Stay updated with the latest news and announcements from i10X.","publisher":{"@id":"https:\/\/news.i10x.ai\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/news.i10x.ai\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/news.i10x.ai\/#organization","name":"i10X","url":"https:\/\/news.i10x.ai\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/news.i10x.ai\/#\/schema\/logo\/image\/","url":"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/10\/i10x-logo-180.png","contentUrl":"https:\/\/i10xnews.kinsta.cloud\/wp-content\/uploads\/2026\/10\/i10x-logo-180.png","width":180,"height":180,"caption":"i10X"},"image":{"@id":"https:\/\/news.i10x.ai\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/i10x_ai","https:\/\/www.linkedin.com\/company\/i10x-ai","https:\/\/instagram.com\/i10x.ai","https:\/\/www.youtube.com\/@i10X_ai_official"]},{"@type":"Person","@id":"https:\/\/news.i10x.ai\/#\/schema\/person\/d410bb7ee72febefca2c9c369939219a","name":"Christopher Ort","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","url":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","contentUrl":"https:\/\/i10x.ai\/news\/wp-content\/uploads\/2026\/10\/christopher-ort-192.webp","caption":"Christopher Ort"},"description":"Christopher Ort handles SEO, AI agent workflows and PR & partnerships at i10x.ai, the all-in-one AI workspace built by i10X Global Pte. Ltd. in Singapore. He is a graduate of Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU). He is the author of \u201cThe Style Penalty\u201d, a study published on SSRN that analyses 1,576 evaluation decisions by four frontier AI models to show how AI-written CVs fare in automated hiring screens; the research has been covered by heise online and t3n.","url":"https:\/\/i10x.ai\/news\/author\/christopher-ort","sameAs":["https:\/\/i10x.ai\/blog\/author\/christopher-ort"],"worksFor":{"@id":"https:\/\/i10x.ai\/news\/#organization"},"alumniOf":{"@type":"EducationalOrganization","name":"Friedrich-Alexander-Universit\u00e4t Erlangen-N\u00fcrnberg (FAU)","url":"https:\/\/www.fau.eu"}}]}},"_links":{"self":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts\/6102","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/comments?post=6102"}],"version-history":[{"count":2,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts\/6102\/revisions"}],"predecessor-version":[{"id":6253,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/posts\/6102\/revisions\/6253"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/media\/6252"}],"wp:attachment":[{"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/media?parent=6102"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/categories?post=6102"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/tags?post=6102"},{"taxonomy":"company","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/companies?post=6102"},{"taxonomy":"product","embeddable":true,"href":"https:\/\/i10x.ai\/news\/wp-json\/wp\/v2\/products?post=6102"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}