{"id":2287,"date":"2026-09-28T08:35:54","date_gmt":"2026-09-28T08:35:54","guid":{"rendered":"https:\/\/newsraise.com\/in\/2026\/09\/28\/ai-advice-reduces-accuracy-raises-confidence\/"},"modified":"2026-09-28T08:35:54","modified_gmt":"2026-09-28T08:35:54","slug":"ai-advice-reduces-accuracy-raises-confidence","status":"publish","type":"post","link":"https:\/\/newsraise.com\/in\/2026\/09\/28\/ai-advice-reduces-accuracy-raises-confidence\/","title":{"rendered":"Study Finds AI Advice Cuts Accuracy While Boosting User Confidence"},"content":{"rendered":"\n<!-- Quick Adsense WordPress Plugin: http:\/\/quickadsense.com\/ -->\n<div class=\"9fece8afa224fd09e54b043d0febfb58\" data-index=\"1\" style=\"float: none; margin:10px 0 10px 0; text-align:center;\">\n<script async src=\"https:\/\/pagead2.googlesyndication.com\/pagead\/js\/adsbygoogle.js\"><\/script>\r\n<!-- NR ATF -->\r\n<ins class=\"adsbygoogle\"\r\n     style=\"display:block\"\r\n     data-ad-client=\"ca-pub-8898941184964366\"\r\n     data-ad-slot=\"4839033563\"\r\n     data-ad-format=\"auto\"\r\n     data-full-width-responsive=\"true\"><\/ins>\r\n<script>\r\n     (adsbygoogle = window.adsbygoogle || []).push({});\r\n<\/script>\n<\/div>\n<p>When an artificial\u2011intelligence chatbot supplies an answer, users may feel assured even if the information is wrong. A recent study involving researchers from three French and Italian universities confirms that the mere availability of AI advice can dramatically erode accuracy and suppress the habit of admitting uncertainty.<\/p>\n<h2>Key findings from the study<\/h2>\n<p>The experiment measured participants\u2019 performance on a set of factual questions under two conditions: with and without access to an AI\u2011generated response. When AI assistance was offered, correct answers fell from 27\u202fpercent to just 9\u202fpercent. At the same time, self\u2011reported confidence surged from 30\u202fpercent to 76\u202fpercent.<\/p>\n<p>Perhaps more striking was the collapse of participants\u2019 willingness to say \u201cI don\u2019t know.\u201d Without AI, 44\u202fpercent of respondents chose that option when unsure; with AI present, the figure dropped to a mere 3\u202fpercent. The researchers concluded that AI tools, which are designed to provide answers rather than acknowledge uncertainty, are teaching users to mimic that same behavior.<\/p>\n<h2>Expert commentary on \u201ccognitive surrender\u201d<\/h2>\n<p>These results echo earlier work from The Wharton School, where scholars coined the term \u201ccognitive surrender.\u201d Wharton researchers observed that people accept incorrect AI answers about 80\u202fpercent of the time and report higher confidence than those who work without AI. Gideon Nave of Wharton explained the distinction: while a calculator offloads a specific arithmetic task but leaves human reasoning intact, cognitive surrender occurs when AI makes a decision and the user adopts it without recognizing the transfer of judgment.<\/p>\n<p>One of the French\u2011Italian study\u2019s authors, Valerio\u202fCapraro, warned that children growing up with such systems may never develop the critical\u2011thinking habits needed to question a machine\u2019s output. The concern extends beyond individual errors; it touches on how AI reshapes collective knowledge practices.<\/p>\n<h2>Implications for internet search and information diversity<\/h2>\n<p>The rise of AI\u2011generated overviews in Google Search illustrates the broader shift toward \u201csingle\u2011front\u2011door\u201d access to information. Jaspreet\u202fBindra, co\u2011founder and CEO of AI\u202f&#038;\u202fBeyond, told ETV Bharat that while AI\u2011curated answers boost convenience, they risk narrowing the range of perspectives users encounter. \u201cThe internet was built on exploration; AI risks turning it into a set of pre\u2011packaged answers unless we consciously design for plurality,\u201d he said.<\/p>\n<p>Abhishek\u202fAgarwal, President of Judge India &#038; Global Delivery, emphasized that the core problem is not occasional AI errors but the habit of users ceasing to notice their own knowledge gaps once an AI answer appears. \u201cWillingness to admit uncertainty collapses almost entirely once an AI answer is available, even as accuracy on the same questions drops sharply,\u201d he noted.<\/p>\n<p>Prior research supports these observations. Studies by teams at Carnegie Mellon University, Oxford, MIT and UCLA found that a brief, ten\u2011minute session of AI assistance left participants less capable of solving math problems. Separate investigations reported reduced brain activity among essay writers, diminished independent cancer\u2011detection ability among doctors, and weakened critical\u2011thinking skills among data workers after relying on AI tools.<\/p>\n<h2>Recommendations for responsible AI use<\/h2>\n<p>Agarwal outlined a practical framework for preserving judgment while still benefiting from AI. He advises users to formulate a rough answer before consulting the tool, forcing an honest self\u2011assessment of uncertainty. The AI\u2019s output should be treated as a \u201csecond opinion worth arguing with,\u201d not a final verdict.<\/p>\n<p>For verification, he suggests a triage approach rather than exhaustive fact\u2011checking. High\u2011stakes decisions\u2014those involving numbers that will be acted upon, public claims, or consequential outcomes\u2014should be rigorously cross\u2011checked. Low\u2011stakes uses, such as brainstorming or drafting, can proceed with lighter scrutiny.<\/p>\n<p>When possible, users can ask the AI to flag its own uncertainty and cite sources, which many models will do if prompted directly. Spot\u2011checking only the claims that would alter a decision if wrong offers an efficient balance between confidence and diligence.<\/p>\n<p>Finally, Agarwal recommends using AI to stress\u2011test conclusions that users have already reached, rather than allowing the system to generate those conclusions outright. \u201cReasoning works like a muscle,\u201d he said. \u201cQuietly outsourcing every rep weakens it over time, even when each shortcut feels harmless in the moment.\u201d Professionals who extract real value from AI, he added, treat the technology as a \u201csharp sparring partner,\u201d not a replacement for independent thought.<\/p>\n<p>As AI becomes an ever\u2011more pervasive front\u2011end to the internet, these findings highlight a paradox: tools designed to augment human capability may inadvertently erode the very cognitive habits that underpin accurate judgment. The challenge for developers, educators and users alike is to embed safeguards that preserve critical thinking while still delivering the convenience that has made large language models so popular.<\/p>\n\n<div style=\"font-size: 0px; height: 0px; line-height: 0px; margin: 0; padding: 0; clear: both;\"><\/div>","protected":false},"excerpt":{"rendered":"<p>A new study shows that when people can consult AI, their answer accuracy falls dramatically while their confidence and willingness to admit uncertainty collapse.<\/p>\n","protected":false},"author":3,"featured_media":2288,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[299],"tags":[231,2478,2480,2479,1804],"class_list":["post-2287","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-artificial-intelligence","tag-cognitive-bias","tag-internet","tag-research","tag-technology","entry"],"_links":{"self":[{"href":"https:\/\/newsraise.com\/in\/wp-json\/wp\/v2\/posts\/2287","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/newsraise.com\/in\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/newsraise.com\/in\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/newsraise.com\/in\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/newsraise.com\/in\/wp-json\/wp\/v2\/comments?post=2287"}],"version-history":[{"count":0,"href":"https:\/\/newsraise.com\/in\/wp-json\/wp\/v2\/posts\/2287\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/newsraise.com\/in\/wp-json\/wp\/v2\/media\/2288"}],"wp:attachment":[{"href":"https:\/\/newsraise.com\/in\/wp-json\/wp\/v2\/media?parent=2287"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/newsraise.com\/in\/wp-json\/wp\/v2\/categories?post=2287"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/newsraise.com\/in\/wp-json\/wp\/v2\/tags?post=2287"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}