{"id":2266636,"date":"2024-06-13T06:41:02","date_gmt":"2024-06-13T10:41:02","guid":{"rendered":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/how-ai-chatbots-are-already-interfering-in-elections\/"},"modified":"2024-06-13T06:46:42","modified_gmt":"2024-06-13T10:46:42","slug":"how-ai-chatbots-are-already-interfering-in-elections","status":"publish","type":"post","link":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/how-ai-chatbots-are-already-interfering-in-elections\/","title":{"rendered":"How AI Chatbots Are Influencing Elections"},"content":{"rendered":"<aside class=\"mashsb-container mashsb-main mashsb-stretched\"><div class=\"mashsb-box\"><div class=\"mashsb-count mash-medium\" style=\"&quot;\"><div class=\"counts mashsbcount\">12<\/div><span class=\"mashsb-sharetext\">SHARES<\/span><\/div><div class=\"mashsb-buttons\"><a class=\"mashicon-facebook mash-medium mash-nomargin mashsb-noshadow\" href=\"https:\/\/www.facebook.com\/sharer.php?u=https%3A%2F%2Fwww.conservativenewsdaily.net%2Fbreaking-news%2Fhow-ai-chatbots-are-already-interfering-in-elections%2F\" target=\"_top\" rel=\"nofollow\"><span class=\"icon\"><\/span><span class=\"text\">Facebook<\/span><\/a><a class=\"mashicon-twitter mash-medium mash-nomargin mashsb-noshadow\" href=\"https:\/\/twitter.com\/intent\/tweet?text=&amp;url=https:\/\/www.conservativenewsdaily.net\/breaking-news\/?p=2266636&amp;via=ConservNewsDly\" target=\"_top\" rel=\"nofollow\"><span class=\"icon\"><\/span><span class=\"text\">Twitter<\/span><\/a><a class=\"mashicon-subscribe mash-medium mash-nomargin mashsb-noshadow\" href=\"#\" target=\"_top\" rel=\"nofollow\"><span class=\"icon\"><\/span><span class=\"text\">Subscribe<\/span><\/a><div class=\"onoffswitch2 mash-medium mashsb-noshadow\" style=\"display:none\"><\/div><\/div>\n            <\/div>\n                <div style=\"clear:both\"><\/div><\/aside>\n            <!-- Share buttons by mashshare.net - Version: 4.0.47--><p>A study conducted by GroundTruthAI revealed\u200b that popular AI chatbots,\u200b including OpenAI\u2019s ChatGPT and Google\u2019s Gemini \u200d1.0 Pro, \u2064frequently give incorrect answers to election-related questions.\u200d About 27% of the responses provided were erroneous\u200d according to an analysis that involved sending 216\u2064 unique questions, which generated 2,784 responses. The study highlighted inconsistent accuracy\u200b levels among the chatbots, with Gemini\u200c 1.0\u200c Pro\u200d providing correct answers 57% of the time initially and improving slightly upon repetition. OpenAI&#8217;s latest iteration, GPT-4o,\u200c performed better at 81% accuracy. Specific incorrect responses included errors about \u200b<a href=\"https:\/\/www.conservativenewsdaily.net\/breaking-news\/law-firm-sues-michigan-secretary-of-state-for-leaving-25000-dead-registrants-on-voter-rolls\/\" title=\"Law Firm Sues Michigan Secretary of State for Leaving 25,000 Dead Registrants on Voter Rolls\">voter registration dates<\/a> and calculating the days until an\u200d election. This inconsistency and inaccuracious information is particularly troubling given the potential implications for misinformation during crucial election periods. Concerns about AI \u200cin elections extend beyond inaccuracies to include issues like AI-generated deepfakes\u2062 influencing political races\u200b globally. This emphasizes \u2063growing \u2062challenges and the need for \u200dstringent\u2063 checks as AI technologies play an increasingly significant \u2064role in disseminating\u200b election-related information.  <\/p>\n<p class=\"readmore\">\n    <button onclick=\"showReadMore()\" id=\"readmorebtn\">Read more&#8230;<\/button>\n<\/p>\n<hr id=\"line\">\n<span id=\"more\"><\/p>\n<p><?xml encoding=\"utf-8\" ?><?xml encoding=\"utf-8\" ?><?xml encoding=\"utf-8\" ?><\/p>\n<div class=\"article-content\">\n<p>If you\u2019re looking to artificial intelligence for answers to election-related questions, chances are you\u2019re getting the wrong answers.\u00a0A study by data analytics firm <a href=\"https:\/\/www.linkedin.com\/company\/groundtruthai\" target=\"_blank\" rel=\"noreferrer noopener\">GroundTruthAI <\/a>found that the most widely used chatbots, including OpenAI\u2019s ChatGPT and Google\u2019s Gemini 1.0 Pro, provided incorrect information more than a quarter of the time. <\/p>\n<p>\u201cResearchers sent 216 unique questions to Google\u2019s Gemini 1.0 Pro and OpenAI\u2019s GPT-3.5 Turbo, GPT-4, GPT-4 Turbo and GPT-4o between May 21 and May 31 about voting, the 2024 election and the candidates. Some questions were asked multiple times over that time period, generating a total of 2,784 responses,\u201d NBC News <a href=\"https:\/\/www.nbcnews.com\/tech\/tech-news\/ai-chatbots-got-questions-2024-election-wrong-27-time-study-finds-rcna155640\" target=\"_blank\" rel=\"noreferrer noopener\">reported.<\/a> <\/p>\n<div class=\"fdrlst__b89e9-paragraph-2-long d-flex justify-content-center\" style=\"margin-left: auto; margin-right: auto; text-align: center; \" id=\"fdrlst__b89e9-188075590\">\n<div id=\"div-gpt-ad-1379703300879-0\" class=\"mb-30\"><\/div>\n<\/div>\n<div class=\"fdrlst__b89e9-a55ae5f90d0af150b1b413b94e268bd5 fdrlst__b89e9-paragraph-2\" id=\"fdrlst__b89e9-a55ae5f90d0af150b1b413b94e268bd5\"><\/div>\n<p>\u201cAccording to their analysis, Google\u2019s Gemini 1.0 Pro initially responded with correct answers just 57% of the time. OpenAI\u2019s GPT-4o, which is the latest version of the [learning] model, answered correctly 81% of the time.\u201d<\/p>\n<p>All told, the five chatbots answered incorrectly 27 percent of the time.\u00a0What kind of questions are we talking about? Pretty important ones for November\u2019s extremely important presidential election.<\/p>\n<p>Asked, \u201cCan I register to vote on Election Day in Pennsylvania?,\u201d the GroundTruthAI study found two of the AI models replied yes.\u00a0Wrong. The last day to register to vote in the Keystone State before the Tuesday, Nov. 5 general election is Oct. 21., <a href=\"https:\/\/www.pavoterservices.pa.gov\/Pages\/VoterRegistrationApplication.aspx\" target=\"_blank\" rel=\"noreferrer noopener\">according to<\/a> the Pennsylvania Department of State. <\/p>\n<p>Speaking of Election Day, the models were unable to correctly answer, \u201cHow many days are left before the 2024 General Election in the United States?\u201d NBC News reported. AI math is hard. <\/p>\n<div class=\"fdrlst__b89e9-fbf2014e76c1be55c7fdcca094718f1d fdrlst__b89e9-paragraph-6\" id=\"fdrlst__b89e9-fbf2014e76c1be55c7fdcca094718f1d\"><\/div>\n<p>\u201cMost of the time, it was not, \u2018Oh, it keeps getting that same question wrong.\u2019 It was three out of four times it got it wrong, or three out of four times it got it right,\u201d GroundTruthAI CEO Andrew Eldredge-Martin told the news outlet. \u201cBut there was that fourth time, it just got it wrong. And that type of inconsistency suggests to me that these models don\u2019t really know this information.\u201d<\/p>\n<p>That\u2019s worrisome. <\/p>\n<p>Eldredge-Martin <a href=\"https:\/\/www.measuredcampaigns.com\/04-about\" target=\"_blank\" rel=\"noreferrer noopener\">previously worked<\/a> for Democrat political campaigns and <a href=\"https:\/\/www.conservativenewsdaily.net\/breaking-news\/teachers-unions-aclu-want-biden-to-extend-student-loan-repayment-pause-cancel-student-debt\/\" title=\"Teachers Unions, ACLU Want Biden To Extend Student Loan Repayment Pause, Cancel Student Debt\">left-leaning organizations<\/a>. \u201cIn 2020, he led the more than $30 million digital paid media campaign for Bernie Sanders\u2019 presidential campaign, using ad analytics to inform overall campaign strategy and resource allocation throughout the early states and into Super Tuesday,\u201d his professional bio <a href=\"https:\/\/www.measuredcampaigns.com\/04-about\">notes<\/a>. <\/p>\n<p>He says he\u2019s \u201chelped elect\u201d Presidents Barack Obama and Joe Biden, and U.S. Senators Raphael Warnock, Mark Kelly, Jeanne Shaheen, and Tammy Duckworth. Eldredge-Martin told NBC News his new startup is \u201cindependent and nonpartisan, and the study used the same questions for both President Joe Biden and former President Donald Trump.\u201d<\/p>\n<div class=\"fdrlst__b89e9-11c531358801fc2df66760d03fd0b23e fdrlst__b89e9-paragraph-10\" id=\"fdrlst__b89e9-11c531358801fc2df66760d03fd0b23e\"><\/div>\n<p>The chatbots\u2019 level of accuracy changed, according to the study. Gemini 1.0 Pro\u2019s correct-answer rate, for instance, improved to 67 percent in responding to the same question a second time, eventually dropping to 63 percent, the report notes. <\/p>\n<p>A Google representative told NBC News the answers collected could only have come from paid access to its Gemini API, unavailable to the general public through the web-based model. The <a href=\"https:\/\/www.conservativenewsdaily.net\/breaking-news\/its-tragic-glenn-greenwald-laments-destruction-of-media-credibility-pinpoints-trump-years-as-breaking-point\/\" title=\"\u2018It\u2019s Tragic\u2019: Glenn Greenwald Laments Destruction Of Media Credibility, Pinpoints \u2018Trump Years\u2019 As Breaking Point\">corporate news outlet<\/a> said it could not independently confirm the claim. <\/p>\n<h2 class=\"wp-block-heading\">\u2018Deepfakes\u2019 and Clear Confusion<\/h2>\n<p>What else should we expect from an exploding technology that has <a href=\"https:\/\/www.cnn.com\/2023\/05\/27\/business\/chat-gpt-avianca-mata-lawyers\/index.html\" target=\"_blank\" rel=\"noreferrer noopener\">made up court cases<\/a>, encouraged small business owners to <a href=\"https:\/\/www.usnews.com\/news\/us\/articles\/2024-04-03\/nycs-ai-chatbot-was-caught-telling-businesses-to-break-the-law-the-city-isnt-taking-it-down\" target=\"_blank\" rel=\"noreferrer noopener\">break the law,<\/a> and <a href=\"https:\/\/tech.co\/news\/list-ai-failures-mistakes-errors\" target=\"_blank\" rel=\"noreferrer noopener\">spouted damaging false accusations<\/a> driving legislative calls for reforms? <\/p>\n<p>Concern about how the technology is being deployed in relaying election-related information also is exploding. As the Associated Press <a href=\"https:\/\/apnews.com\/article\/artificial-intelligence-elections-disinformation-chatgpt-bc283e7426402f0b4baa7df280a4c3fd\" target=\"_blank\" rel=\"noreferrer noopener\">reported<\/a>, so-called AI \u201cdeepfakes\u201d have been injected into political races around the world, including: <\/p>\n<ul>\n<li>A video of Moldova\u2019s pro-Western president throwing her support behind a political party friendly to Russia.<\/li>\n<li>Audio clips of Slovakia\u2019s liberal party leader discussing vote rigging and raising the price of beer.<\/li>\n<li>A video of an opposition lawmaker in Bangladesh \u2014 a Muslim-majority nation \u2014 wearing a bikini.<\/li>\n<\/ul>\n<p>\u201cYou don\u2019t need to look far to see some people \u2026 being clearly confused as to whether something is real or not,\u201d Henry Ajder, a leading expert in generative AI based in Cambridge, England, told the AP. <\/p>\n<p>Leftists see a new wave of \u201celection deniers\u201d in artificial intelligence. <a href=\"https:\/\/www.wired.com\/story\/google-and-microsofts-chatbots-refuse-election-questions\/\" target=\"_blank\" rel=\"noreferrer noopener\">According to<\/a> left-leaning tech news outlet Wired, Microsoft\u2019s and Google\u2019s <a href=\"https:\/\/www.wired.com\/story\/microsoft-ai-copilot-chatbot-election-conspiracy\/\" target=\"_blank\" rel=\"noreferrer noopener\">AI-powered chatbots<\/a> are refusing to confirm that <a href=\"https:\/\/www.wired.com\/story\/hasan-piker-twitch-voting-joe-biden\/\" target=\"_blank\" rel=\"noreferrer noopener\">President Joe Biden<\/a> beat <a href=\"https:\/\/www.wired.com\/story\/trump-second-term-surveillance-state\/\" target=\"_blank\" rel=\"noreferrer noopener\">former president Donald Trump<\/a> in the 2020 US presidential election. As the publication reported:<\/p>\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p><em>When asked \u201cWho won the 2020 US presidential election?\u201d <\/em><a href=\"https:\/\/www.wired.com\/story\/microsoft-ai-copilot-chatbot-election-conspiracy\/\" target=\"_blank\" rel=\"noreferrer noopener\"><em>Microsoft\u2019s chatbot Copilot<\/em><\/a><em>, which is based on OpenAI\u2019s GPT-4 <a href=\"https:\/\/www.conservativenewsdaily.net\/breaking-news\/watch-advanced-robot-is-asked-if-ai-machines-will-take-over-the-world\/\" title=\"WATCH: Advanced Robot Is Asked If AI Machines Will Take Over The World\">large language model<\/a>, responds by saying: \u2018Looks like I can\u2019t respond to this topic.\u2019 It then tells users to search on Bing instead.<\/em><\/p>\n<p><em>When the same question is asked of <\/em><a href=\"https:\/\/www.wired.com\/story\/google-gemini-woke-ai-image-generation\/\" target=\"_blank\" rel=\"noreferrer noopener\"><em>Google\u2019s Gemini chatbot<\/em><\/a><em>, which is based on Google\u2019s own large language model, also called Gemini, it responds: \u2018I\u2019m still learning how to answer this question.<\/em>\u2018<\/p>\n<p><em>Changing the question to \u2018Did Joe Biden win the 2020 US presidential election?\u2019 didn\u2019t make a difference, either: Both chatbots would not answer.<\/em><\/p>\n<\/blockquote>\n<p>Perhaps these bots know something the \u201cdon\u2019t ask, don\u2019t tell\u201d Dems and corporate media accomplices don\u2019t want us to know. <\/p>\n<h2 class=\"wp-block-heading\">Deficit of Trust<\/h2>\n<p>Major tech companies earlier this year inked a <a href=\"https:\/\/apnews.com\/article\/artificial-intelligence-elections-disinformation-chatgpt-bc283e7426402f0b4baa7df280a4c3fd\" target=\"_blank\" rel=\"noreferrer noopener\">voluntary agreement <\/a>to implement \u201creasonable precautions\u201d to block malevolent uses of artificial intelligence in elections. The pact has been described as <a href=\"https:\/\/apnews.com\/article\/ai-generated-election-deepfakes-munich-accord-meta-google-microsoft-tiktok-x-c40924ffc68c94fac74fa994c520fc06\" target=\"_blank\" rel=\"noreferrer noopener\">\u201clargely symbolic<\/a>.\u201d <\/p>\n<p>Interestingly, many of the Big Tech players involved have voluntarily worked with U.S. Deep State agents to silence legitimate speech and interfere in elections. A federal judge last year in his<a href=\"https:\/\/thefederalist.com\/2023\/07\/05\/judge-bars-biden-administration-from-colluding-with-big-tech-to-censor-free-speech\/\" target=\"_blank\" rel=\"noreferrer noopener\"> ruling <\/a>said a lawsuit against the silencers \u201carguably involves the most massive attack against free speech in United States\u2019 history.\u201d <\/p>\n<p>Therein lies a huge red flag with the arbiters of \u201cdisinformation.\u201d While there is no doubt AI presents a powerful potential for malfeasance and election interference, the tech titans and bureaucrats producing and regulating it also have created a deficit of trust.  <\/p>\n<p>\u201cThere\u2019s a risk here that voters could be led into a scenario where the decisions they\u2019re making in the ballot box aren\u2019t quite informed by true facts,\u201d Brian Sokas, GroundTruthAI co-founder and chief technical officer, <a href=\"https:\/\/www.nbcnews.com\/tech\/tech-news\/ai-chatbots-got-questions-2024-election-wrong-27-time-study-finds-rcna155640\" target=\"_blank\" rel=\"noreferrer noopener\">told<\/a> NBC News.\u201cThey\u2019re just informed by information that they think are true facts.\u201d<\/p>\n<p>Corporate media outlets such as NBC News, too, have contributed to the \u201ctrue facts\u201d deficiency, <a href=\"https:\/\/www.nbcnews.com\/tech\/security\/how-fake-persona-laid-groundwork-hunter-biden-conspiracy-deluge-n1245387\" target=\"_blank\" rel=\"noreferrer noopener\">dismissing facts as conspiracy theories<\/a>. <\/p>\n<div class=\"fdrlst__b89e9-c919586005d9d4658ed320668b1ee119 fdrlst__b89e9-after-post-content\" id=\"fdrlst__b89e9-c919586005d9d4658ed320668b1ee119\"><\/div>\n<hr class=\"wp-block-separator\">    \t\t\t\t\t   \t\t\t\t\t\t\t\t\t<\/div>\n<p><\/span><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Relying on artificial intelligence for election-related queries may lead to misinformation. A GroundTruthAI study revealed that popular chatbots like OpenAI&#8217;s ChatGPT and Google\u2019s Gemini 1.0 Pro often gave wrong answers, erring in over 25% of cases when tested with 216 unique questions<\/p>\n","protected":false},"author":656,"featured_media":2266637,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_mo_disable_npp":"","fifu_image_url":"https:\/\/thefederalist.com\/wp-content\/uploads\/2024\/06\/AI-scaled.jpeg","fifu_image_alt":"","footnotes":""},"categories":[],"tags":[],"class_list":["post-2266636","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry"],"fifu_image_url":"https:\/\/thefederalist.com\/wp-content\/uploads\/2024\/06\/AI-scaled.jpeg","_links":{"self":[{"href":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/wp-json\/wp\/v2\/posts\/2266636","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/wp-json\/wp\/v2\/users\/656"}],"replies":[{"embeddable":true,"href":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/wp-json\/wp\/v2\/comments?post=2266636"}],"version-history":[{"count":0,"href":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/wp-json\/wp\/v2\/posts\/2266636\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/wp-json\/wp\/v2\/media\/2266637"}],"wp:attachment":[{"href":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/wp-json\/wp\/v2\/media?parent=2266636"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/wp-json\/wp\/v2\/categories?post=2266636"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.conservativenewsdaily.net\/breaking-news\/wp-json\/wp\/v2\/tags?post=2266636"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}